ML072710562

Caoldon Non-Proprietary Information Package for Seabrook/Nrc Meeting on December 16, 2005

ML072710562
Person / Time
Site:	Seabrook
Issue date:	12/16/2005
From:	Caldon
To:	Office of Nuclear Reactor Regulation
References
FOIA/PA-2007-0255
Download: ML072710562 (173)

Text

S

~

0 0

Caldon Non-Proprietary Information Package for Seabrook/NRC Meeting December 16, 2005 INFO-19

Non-Proprietary Information Package for Seabrook/NRC December 16,12005 INFO TABLE OF CONTENTS

1. Profile Factor Reports Listing

2. TP77 - Traceability of Thermal Power Measurements, Part 2 - Modified Venturi Tubes

3. TP55 - Effects Of Varying Hydraulics On The Calibration Of Eight Path Chordal Ultrasonic Meters

4. TP44 - Theory Of Ultrasonic Flow Measurement-Gases And Liquids

5. ER-262 - Effects Of Velocity Profile Changes Measured In-Plant On Feedwater Flow Measurement Systems

6. PID's, Reference Drawings, and ISO's

7. TP76 - Traceability of Thermal Power Measurements, Part I - Chordal Ultrasonic Flow Measurements

8. CNUG Agendas and Attendance Sheets

9. Example of transmittal letter with Profile Factor Final Report - PB 10-17-03 LEFM CofC on System INFO-19

Profile Factor Calculations ER No.

Date Description Description Date of Test Check and CheckPlus Meters Alden Report Number ER-265 RO 12/20/01 ASCO 1&2 ARL NO.356-01/C730 ER-175 RO 9/11/00 Beaver Valley ARL NO.279-00/C730 ER-254 RO 10/19/01 Cofrentes ARL NO.340-01/C730 ER-227 RI 10/08/01 Davis Besse ARLNO.310-01/C730 ER-287 RI 4/22/02 DC Cook 1 ARLNO.121-02/C730 ER-320 RO 1/20/03 DC Cook 2 ARLNO.23-03/C730 ER-182 RO 12/13/00 Grand Gulf ARL NO.388-00/C730 ER-295 R2 06/10/02 H5 ARL NO.170-02/C730 ER-406 RO 12/10/03 Ikata 1 ARL NO.324-03/C730 ER-407 RO 12/17/03 Ikata 2 ARL NO.325-03/C730 ER-416 RO 02/11/04 Ikata 3 ARL NO. 35-04/C730 ER-394 RO 10/09/03 Millstone 3 ARL NO.248-03/C702 ER-292 RO 5/21/02 Peach Bottom 2 ARL NO. 148-02/C730 ER-441 RO 6/30/04 Peach Bottom 2 ARL NO. 147-04/C730 ER-375 RO 6/23/03 Peach Bottom 3 ARL NO.141-03/C730 ER-327 RI 1/23/03 River Bend ARL NO.29-03/C730 ER-300 RO 7/08/02 Robinson ARL NO. 212-02/C730 ER426 R1 2/25/04 Robinson ARL NO. 44-02/C730 ER-223 R2 09/05/01 Sequoyah 1 ARL NO. 262-01/C730 ER-277 RO 09/19/01 Sequoyah 2 ARL NO.284-01/C730 ER-219 RO 7/30/01 Susquehanna 1 ARL NO. 241-01/C730 ER-199 RO 1/05/01 Susquehanna 2 ARL NO. 01-01/C730 ER-264 R2 12/14/01 Vandellos 2 ARL NO.355-01/C730

Profile Fac-tor Calculations ER No.

Date Description Description Date of Test Check and CheckPlus Meters Alden Report Number ER-214 RO 06/19/01 Waterford ARL NO.195-01/C730 ER-168 R2 8/15/00 Watts Bar ARL NO.252-00/C730

Traceability of Thermal Power Measurements Don Augenstein Herb Estrada Ernie Hauser Caldon, Inc.

1070 Banksville Avenue Pittsburgh, PA 15216 daugenstein( caldon.net herbestrada(comcast.net ehauser(ecaldon.net Keywords: traceability, chordal, transit-time, ultrasonic, measurement, venturi, modified venturi Part 2 Modified Venturi Tubes ABSTRACT This is the second of two papers describing the traceability of nuclear feedwater flow measurements. The first considered the challenges and methodology for establishing the traceability of chordal ultrasonic fldw meters. This paper considers the challenges of establishing the traceability in a measurement using a flow element of the modified venturi tube type. It specifically considers the. assumptions and uncertainties associated with the extrapolation, for use in the field, of tube calibration factors measured -in the laboratory. To quantify these uncertainties, the in-situ performance of four modified venturi tubes is compared with the performance of four 8-path chordal ultrasonic flowmeters. The data analyzed were collected in the feeds of four steam generators in a large pressurized water reactor plant, each feed containing one meter of each type. The meters were initially calibrated in this series arrangement in a NIST traceable calibration lab and then operated in the same arrangement in the field.

1. INTRODUCTION A continuous, accurate determination of thermal power is essential in the operation of a nuclear power plant. Errors in the power determination can cause lost revenue or reduced safety margin-both serious consequences. It is therefore appropriate that the rigor of traceability be applied to each component of the thermal power determination. The key element in the determination of thermal power is the measurement of the mass rate of feedwater flow. The desirability of applying rigorous traceability requirements to the feedwater flow measurement is underlined by recent problems with flow instrumentation in nuclear applications.

TP77 Rev. L.doc

Traceability is defined as a process whereby a measurement can be related to a standard via a chain of comparisons (International Standards Organization, (1)). The companion to this paper (Augenstein, et al, 2)) listed the following key elements of traceability:

The standard must be acceptable to all parties with an interest in the measurement and is usually a standard maintained by a national laboratory such as the National Institute of Standards and Technology.

• The chain of comparisons must be unbroken--the field measurement must be connected, by one or more links directly to the standard.

• Every link in the chain involves a comparison that necessarily carries with it an uncertainty. Hence the total uncertainty of the measurement must reflect the aggregate uncertainties of each link of the comparison chain.

• There can be no unverified assumptions in the chain of comparisons; it is clearly not possible rationally to assign an uncertainty to an assumption with no quantitative basis.

This paper analyzes the traceability chains for flow elements of the modified venturi tube type', from their basic measurements--the differential pressure between the upstream and throat taps of the modified venturi and the fluid temperature and pressure--to the process variable, feedwater mass flow. Much of the discussion applies qualitatively to nozzle-type flow elements. The paper covers explicitly the calibration uncertainties of the flow element(s), including the application of the flow element calibration data taken in a hydraulics facility operating at 100 F and 50 psig to the 430 to 450 F, 1000 to 1200 psig conditions in a nuclear feedwater system at full power.

2. DISCUSSION The Algorithm for a Nozzle or Venturi -Based Mass Flow Measurement The algorithm for the determination of mass flow of water for a differential producer (nozzle, Venturi, orifice, etc) is as follows (refer to the Fluid Meters, (4)):

d2 q*, =0.0997019C d

1 41-(d/D))

Where q*,,,, is the mass flow rate in lbs/s C discharge coefficient (dimensionless) d throat diameter (inches at the flowing temperature) d = 1 + crh,. (Tfo~g-680 F)d68oF Where d6 9, F is the throat diameter at680F

+ The modified Venturies whose calibrations are described in this paper are similar to the "Universal Tube" (trademark, General Signal Corp.) described by Halmi, (3)

TP77 Rev. I (2)

D Diameter of inlet (inches at the flowing temperature)

D = 1+ a,,

-T7., 680 F)D6 8ROF D6.,. Inlet Venturi diameter at 680F h..69,F Differential pressure in inches of water at 680F p

Fluid density in Ibm/cubic ft a*,,,

Thermal expansion factor of inlet Venturi section ayrt Thermal expansion factor of throat material The fluid density is a function of both pressure and temperature

.p = fp (T, p)

Here T is the temperature of the feedwater.

The function fp can be defined with high accuracy using the equations for water in an appropriate table (e.g., the ASME Steam Tables, (5))

Note: In many older installations the thermal expansion factor is incorrect by 0.1 to 0.2 %

if the upstream meter section is of another material -than the throat.

Elements of the Traceability and Accuracy of a Nozzle or Venturi Tube Based Flow Measurement Fundamentally, the traceability of the mass flow algorithm for a venturi or nozzle type meter requires that a chain of comparisons be constructed for the following elements of the algorithm described above:

• The Discharge Coefficient, C i

The Fluid Temperature Measurement T

• The Differential Pressure Measurement h

• The Fluid Pressure Measurement p This listing presumes that the nozzle or venturi tube has been calibrated in a certified facility, to establish a discharge coefficient, C. The calibration embeds any errors in the measurement of the throat diameter, d or the upstream diameter D. It also presumes that the uncertainties in the function fp are established on a one-time basis by reference to appropriate standards and are small relative to other uncertainties in the flow measurement.

The principal challenges and uncertainties in the flow measurement are associated with

• the discharge coefficient and differential pressure instrument. The density of compressed water is very insensitive to fluid pressure. Reasonable care in the installation, calibration and maintenance of a Resistance Temperature Detector (an RTD), typically employed for feedwater temperature measurements can yield a traceable accuracy in the +/-I-'1F range, which translates to an uncertainty in the mass flow measurement of less than -0.05%.

TP77 Rev. I (3)

Modem instrumentation and digital signal processing removes many of the uncertainties associated with the computation of flow from a differential pressure measurement, particularly in the performance of the multiplications and the square root function. If care is taken in the selection of the instrument range, high quality transmitters are used, attention is paid to the arrangement of impulse lines, and calibrations are performed with high quality, traceable test equipment on a periodic basis, the differential pressure measurement can contribute an uncertainty of no more than +/-0.5% to -0.75% of rated flow.

The key determinant in the accuracy in feedwater flow measurements with venturis or nozzles is the first of the list of traceable elements above-the determination of the discharge coefficient in an appropriate calibration facility and the extrapolation of the.

coefficient thus determined to the field, where fluid conditions-specifically viscosity-will be very different from those of the facility. Quantifying the uncertainties of the facility measurements is straightforward; there are the uncertainties of the facility standards-the weigh tank, the time measurements, the fluid temperature and pressure measurements, and the secondary standard used for the differential pressure measurement-and the calibration technique itself-the repeatability of the diverter mechanism, etc. It is the extrapolation of the discharge coefficient thus determined for use in the field that presents the challenges, particularly, assumptions that are difficult to verify in quantitative terms. Specifically:

" The discharge coefficient is sensitive to global fluid velocity fields. Both the axial and transverse fluid velocities will differ in some degree from lab to field. The sensitivity of nozzles and s to axial velocity profile is usually small but not negligible (Halmi, (6), Ferron, (7)); differences in transverse velocity-in particular swirl--can produce large biases (Fluid Meters, previously cited).

• The discharge coefficient is sensitive to Reynolds Number. The boundary layer thins as the Reynolds Number increases from lab conditions (1 to 3 x 10 ) to field conditions ( to 3 x 107). For some specific nozzle designs there have been theoretical treatments of the impact of the thinning on Discharge Coefficient (e.g.,

Benedict, (8)) but experimental proof of these analyses in Reynolds Number regime for 450 F feedwater has been very limited.

• The sensitivity of discharge coefficient to Reynolds Number is not limited to the thinning of the boundary layer. Separation and reattachment effects are also sensitive to Reynolds Number (Miller, (9)). Separation "bubbles" (i.e., vortices) can change the form of the velocity field in the throat of a nozzle or venturi and, depending on their location, can cause biases of 1% or more in either direction.

" The deposition of corrosion deposits in the throat of nozzles and venturis

("fouling") can cause a change in the effective internal diameter of the throat (d),

thereby changing the discharge coefficient from the extrapolated value determined at the time of calibration. The deposition is electrochemical in nature and preferentially occurs at the reduced pH attending high power operation*.

In Pressurized Water Reactor plants feedwater is usually treated with a volatile agent such that its pH at room temperature is in the 9.5 range. However, the solubility of the HW and OFH ions is such that, at operating temperature, the pH is reduced to the 7 range. (Estrada, (10)). The change in pH can change the TP77 Rev. 1 (4)

Hence it may not be possible to ensure the cleanliness of a nozzle during the full power run following a shutdown, even when the nozzle is cleaned during that shutdown. The deposition often occurs as full power is approached.

The discharge coefficients of nozzles and venturis are sensitive to the local flow field in the vicinity of the pressure taps, particularly the throat tap. Small upsets in the surface at or near the taps can cause stagnation of the local velocity upstream of the upset. Depending on its location, an upset can cause a high or low bias in the indication of the instrument. The presence of an upset can often be detected in the calibration process, allowing for its correction (by careful smoothing of the surface in the vicinity of the taps). However, the deposition of corrosion products can also create local upsets in the throat surface, as can the cleaning of nozzles with high-energy water jets.

Each of the effects described above requires an assumption regarding the performance of the nozzle or venturi at full power feedwater conditions. The bounding of the uncertainties associated with these assumptions represents the greatest difficulty in establishing an accurate discharge coefficient for venturis and nozzles in feedwater service.

Laboratory Calibration of the Modified Venturi Tubes and the 8-path Chordal UFMS Both the 8 path chordal ultrasonic flowmeters and the modified venturi tubes Whose performance is described in this paper were calibrated in a hydraulic model that simulated the field application of the instruments. With respect to nozzles and venturis, this process is unusual-normally, these devices are calibrated in straight pipe with a flow conditioner at a distanceof about 20 diameters to eliminate any transverse velocity components in the calibration flow field. Field installations are typically 10 to 20 diameters downstream of the closest bend, based on the (unverified) assumption that this distance is sufficient to eliminate flow field disturbances produced by this feature and features further upstream. The approach taken for the modified venturis of this paper reduces the uncertainties in discharge coefficient due to the global flow field in the plant by modeling the features that produce that flow field.

Figure 1 is an artist's sketch of the actual plant installation. Note that the modified venturis are roughly 30 diameters downstream of the header, while the 8-path chordal UFMs are at distances ranging from about 15 to 22 diameters. The varying locations for the UFMs provide access for removal of transducers from individual flow elements.

Figure 2 is a photograph of the hydraulic model used in the calibration tests. The UFM is in the foreground of the photo; the modified venturi, downstream of the UFM, is not visible. As can be seen from the photo, a single steam generator feed was used in the calibration laboratory model. The varying distances of the UFMs were however explicitly modeled in the tests for each instrument package. The effect of installation hydraulics on the flow fields of the individual steam generator feeds was investigated by varying the fraction of the total flow from the individual feeds to the header. Variations in the sign of the electrostatic forces between the throat surface and colloidal corrosion products in the flowing feed such that colloids that were repelled at room temperature are attracted at operating temperature.

TP77 Rev. I (5)

fractions of total feed to the header from the individual supplies were also used to test the sensitivity of meter calibrations to variations in velocity fields. Additionally, straight pipe calibration tests were run for two of the UFM-modified venturi tube packages, as benchmarks.

Figure 1 Arrangement of Chordal UFMs and Modified Venturi Tubes in Plant bill Figure 2 Calibration Arrangement for Modified Venturi Tubes and Chordal UFMs in Certified ydraulics Labora Certified H dranli Lort TP77 Rev. 1 (6)

Calibration of the Chordal Ultrasonic Meters and Extrapolation of the Results to Plant Conditions The companion to this paper (Augenstein, et al, cited previously) describes the methodology whereby the calibration factor of chordal UFMs can be extrapolated from the calibration lab to field conditions with bounded and modest uncertainty. The methodology involves characterizing the meter factor of the chordal system using the "flatness" of the axial velocity profile as measured in a model simulating the hydraulics of the field installation. The model is also varied parametrically to determine the sensitivity of the calibration to changes in flatness. Flatness is defined as the ratio of the axial fluid velocity averaged along the outer (short) chords of the UFM to the axial fluid velocity averaged along the inner (long) chords of the UFM.

Figures 3A, 3B, 3C, and 3D show the meter factors (also called the profile factor, PF) for the UFM flow elements for Loops A, B, C, and D. (In the plant the loops are actually numbered 1, 2, 3 and 4). The meter factors are plotted against the flatness. Also shown on each figure is the theoretical sensitivity of the meter factor to flatness. It will be noted that the calibration data for all UFMs lie within :L0. 15% or less, of the theoretical sensitivity over the range of flatness ratios produced by parametrically varying the flow model, with the modest exception of one set of data for Loop D, Figure 3D (the 0- 50 -50 flow splits from the first, second and third feeds), where the difference approaches 0.2%.

Comparison of the figures also shows that the profiles for Loops A and B are rounder than those for Loops C and D (that is, the flatness ratios for A and B tend to be lower than those for C or D), because, as Figure 1 shows, the latter instruments are closer to the header. Downstream of a sudden contraction a profile is flat, becoming rounder as the profile develops (Schlichting, (11)).

Figures 3A through 3D also show the meter factors implemented on the basis of the calibration data. These are plotted (in green) for the flatness ratios measured in the plant during power escalation from roughly 30% to 100% of rating. It will be noted that the extrapolations of the meter factor for Loops B and D are relatively modest in terms of flatness -a change of 0.02 or less. The extrapolations for Loops A and C are slightly larger (0.02 to 0.04). The flatter-than-predicted profiles are due to the presence of larger swirl in the plant than was present in the lab (swirl is measured from the differences in the measured velocities of outer acoustic paths at the same chordal location).

Nevertheless, the difference in flatness does not lead to a significant extrapolation uncertainty. Specifically the uncertainty in meter factor due to the uncertainty in the hydraulics of the model and the extrapolation fit is accounted at :A-0.14%. This figure-the differential uncertainty of the UFM-also accounts the time measurement uncertainty of the instruments used in the calibration tests and the observational uncertainty-turbulence and other effects cause statistical variation in measured results. It does not include the uncertainty of the facility, which for this test was bounded at -0.15%. The reason it does not is that any facility bias that is present in the calibration of a UFM will also be present, with exactly the same magnitude and sign, in the meter factors for the modified venturi tubes. Since the objective of this paper is to investigate the extrapolation of modified venturi tube meter factors to plant conditions by comparing their indications TP77 Rev. 1 (7)

with the UFMs, the uncertainty associated with the facility itself need not and should be included in the analysis.

Figure 3A Loop A LEFM ChackPlu, CalIbration vs. Flatness 1.0150 1.0100 1.0050

• - 1.0ooo 9

0.9900, 0.9850 0.800 0.820 0.540 01880 0.880 0.900 0.920 0.940 0.95C Flatness Figure 3B Loop B LEFM CheckPlus Calibration vs. Flatness 1.0150 1.0100

1.

%"x 1.0000 0.9950 8tfBi9m POO Is 040-100 10D0-0-0

• -Thenticfal Snitivty

£*raigta pipe 33-0-87 0.0-100 x 1000.&0 ffImplemerled PF W'

0.99M0 0.9850 4--

0.800 0.820 0.840 0.860 0.880 0200 0.920 0.940 Flatbam 0.960 TP77 Rev. I (8)

Figure 3C Loop C LEFM ChackPlus Calibration vs. Flatness 1.0150 1.0100 I.0050 4-4 4........

w.

4

• w I _UUUU m

0-0-100 50-0-50 25-25-50 Theoretical Sensltvity I-- mplemiented PF Ou-0,9950 0.99(30 0.9850 0.9950 0.9900 0.985 0.800 D.A20 0.840 0.860 0.880 0.900 0.920 0.940 0.960 Flatness Figure 3D Loop D LEFM CheckPlus Calibration vs. Flatness U.A.

1.0100 1..000'".-_......

0.9950 0.9850

• 0-50-50

• 0-100-0 0-0-100

.c 25-25-50 4-

• Theoredcal Senstivity p45-rpemented PF 0.800 0.820 0.840 0.860 0.880 0.900 0.920 0.940 0.960 Flatness Ratio TP77 Rev. I (9)

As noted in the introductory discussion, Reynolds Number is a key descriptor for the behavior of venturis and nozzles, since it is a significant factor in the thickness of the boundary layer in the throat of the nozzle. It is therefore appropriate to characterize the response of the meter factors of the chordal UFMs to Reynolds Number, so that a comparison in response can be readily carried out. Figures 4A, 4B, 4C, and 4D plot the meter factor data from the calibration tests against Reynolds Number (based on pipe diameter).

The figures also show the theoretical sensitivity of meter factor to Reynolds Number. The sensitivity arises because increasing Reynolds Number thins the boundary layer and flattens the profile, slightly changing the bias associated with the numerical integration of the four chordal velocity measurements. It will be noted that the data for all UFMs closely follow the theoretical sensitivity (within about 10. 1% to +-0.2%).

The figures also show the meter factors implemented in the field (shown in green) plotted against the range of Reynolds Numbers actually experienced during power escalation from 30% to 100%. Note that there is overlap between lab Reynolds Numbers and plant Reynolds Numbers for all meters, giving high.confidence in the use of the UFMs as comparative standards for the modified venturi tubes in the plant Reynolds Number regime.

Figure 4A Loop A LEFM CheckPlus Calibration Profile Factor vs. Reynolds Number a

lb 2

5.

0.1 1.0 10.0 100.0 Reynolds Number ixsle) 100 Percent LUnel -Cl 147C a

100 Percent Una 3-Cal 147B A

StiraIghl Pipe-Ca1147A Theoreical Sensl["ty to RN it Plant data Log. rrheoretical Sensitivity to RN)

Log. (Plant data)

TP77 Rev. 1 (0o)

Figure 4B Loop B LEFM CheckPlus Calibration Profila Factor vs. Reynolds Number 1.015 0I U-0.1 1.0 10.0 100.0 Reynolds Noumber (xI) 0 Staigt*

Pi-Ca 146A 213 Line 3 and 1/3 Une I-Cal 1489 4

100 Line 3-Ca 14sC o

100 Percent Line 1-Ca 148D Log (Theoretical Serltlvty to RN -*--PlI Data Figure 4C Loop C LEFM ChackPlus Calibration vs. Reynolds Number 0I It.

AP I

11I1 I11 t

l I

I 11 11 I

11 1 I

I I

I t!I I

1 1I1 I

I I

I I I I I

I I

I I I I I I I

lII t,!1 I

I I

I I I I

I 1I 1I1 I I

I I

I I

I I I

I 11I I I I I I

I I

lI l t I I

I I

1 1I I I I

I 1 I 1 1 I I I I

I I

I I I 0

I I

I I I II I

I. I I

11 i 1 I

I 1 I t1 1 I

I I 11 1 1 II 1

1 1

I I

1 1

1 I I I 11 I1 I

I i

ll I

I 1

1 1

I I I II1 1

I 1 11 I 11 1

I I I II1 1

I I111 I

1 1 I I 11 1

1 I

I l

lI I

I I

I I

I I

I I III I

III I

I 11II I

I I

I 11 11 I

I I.1 1

i I I

15 1111 I

i I

i I

I 11 l

1i I" I I

I I1 1

S I 1I 1111 I II I

II I II I

I I

I I I I I S

I I

I li lt,/ "

I I

I I

I I

11 I

I I

I I I I I

11 1

I I

I I1 1 I 1

1 III11 1

I I

11 1

It 11 1

III11 1

I I

II111 1

1 1

Ifl111 1

I I

)1 1

I I

1 1 I

1 111 I

I I1 1

I 11 1

I I I1 I1111I I

I I

I u I I I

I I

I 1 I l 0.80I 1

1 1 II I

I 11 1

I II 11 1

25-25-M0 50-0-50 0-0-10o Theoretcal Sanaltivty

-0 Implemetted PF 0.1000 1.0000 10.0000 100.0000 Reynolds Numberil,0POOP (11)

TP77 Rev. 1

Figure 4D Loop 0 LEFM ChecrcPlus Callbratlon S

1.0050U I

I h

u* lal!

I

£ I

h u ll I

I I

I I I I

I t u I $1 I

III1 i

I 1 I I I I

I I

11I I Il I

I I I II I

I lI tI I

I 1

i I 1i II I

i I I1 I 1 ll I

I 1 1I I I

I I

I I

I I

I I 1111I I

I T Ill 1.0100 I

I tit ill I

t tl I

I 1

1 I

it th III I

I I

t I

II ih I

iI I

11 1I I

~

I I 1

1 1 II I

ttItil 1

1 I

I till I

I I

11111 I

I 0.9000 I.9&5I I

0 I I i I

i I

I I

I I

I I tIitli I

I 1

I ll IlII t

I I

i I I I

I I

II l

I I 1 III I

I1 I I I I

I I

I~

ti tIII I

t*

I I

h IlI I

I I

III I

I i t l

f tI

,I r~l h

l J l

I I

It till 1.0000

~~111 I I I

"ii I

I it Ill I

l l

I I I

"l*,'1 I

I I

I 11 I

I I I I

t Ill I

I I

I 1 1 I

i I

t l

I It 11 f

I I

It utu I

I I

l t

0.9 0 I

I I

I Il I I

'1 I. I IlII I

I I

t l

I I

I I I S Itu I

I I

h i

t l

I I

I 1 1 1

I I

I I 1 11 I1 I

I.

II l1ilt I

I I

I I I II I

I I

I 1 111 I

I I

I lI lt II 11 1

I I

i tIlt I

I' I I III 1

I I

I I

I I I

I II 11 1 I I

I l lI I

I I

I tI lt I

I I

III1 I

Ii tI lt I

I I

I I I 1 II I

I h I II I

I tI I

I I t

I111 I

I I

tIlt I I I

I I I

11 1

I I

It It I II11 1

I I I

it I

III111 II II 50I t

tIlt I

I t

ItII ItI 0-50-50 0-100-0 100/

0-50-50 e-hwefecal msi/ty to RN Implemented 0.1000 1.0000 10.0000 10.00000 Reynolds Numbr/tl.,00.000.

Calibration of the Modified Venturi Tubes and Extrapolation of Results to Plant Conditions; Comparison with Chordal UFMs as Standards Calibration data for the modified venturi tubes are plotted against pipe Reynolds Number in Figures 5A, B, C, and D, for loops A, B, C, and D. [The throat Reynolds Number is normally used to characterize venturis and nozzles but, for ease of comparison with Figures 4, pipe Reynolds Number is used here. For these modified venturis the difference in the two Reynolds Numbers is a constant factor of about 2, throat higher.] It should be pointed out that the highest calibration Reynolds Numbers for the modified venturis is somewhat lower than that for the UFMs. The reason is that, in the calibration facility, cavitation in the throats of the venturis occurred as the Reynolds Number approached 3 million, whereas cavitation does not begin in the UFMs until a Reynolds Number above 3.6 million is reached. [At high flow rates in the calibration facility, the pressure at the UFM, just upstream of the modified venturi tubes, was in the 12 to 15 psig range. In the throat of the venturi, where the fluid velocity is increased by a factor of roughly 4, the attendant reduction in static pressure causes cavitation.]

Also shown in the figures are two empirical bases commonly used to extrapolate the discharge coefficient from the lab to operational Reynolds Numbers. The first (higher) basis for extrapolation is a simple log linear fit of the data; the meter factor is assumed to correlate linearly with the logarithm of the Reynolds Number. There is no theoretical basis for the use of the log linear fit but it seems generally to correlate data for some nozzles well and it appears to do so here. The second basis for extrapolation is the so-TP77 Rev. 1 (12)

called reciprocal square root fit. Here the correlation for discharge coefficient is assumed to be of the form:

Co = a - m/SQRT (RN),

Where RN is Reynolds Number.

This form has some theoretical basis in that it assumes that, as the boundary layer in the throat diminishes in thickness with increasing Reynolds Number, the discharge coefficient approaches an asymptote. For some applications, the asymptote, a, is assumed to be 1, which implies that the discharge coefficient approaches that for a reversible gradual contraction with increasing Reynolds. Number. This form is not used here, however, because the residual error for this form is quite large. Instead the asymptote, a, and the slope, m, are selected to minimize the root mean square error of the data relative to the fit. It will be observed that there is a significant difference in the predictions of the two approaches to extrapolation (1/2 to 1% in CD) as the plant Reynolds Number regime (10 to 20 million) is approached.

The figures also show the discharge coefficients implemented for each modified venturi plotted against the range of Reynolds Numbers they see in the escalation of plant power from 30 to 100%. The coefficients were chosen on the basis of the reciprocal square root fit of the data.

The aggregate differential uncertainty in the discharge coefficients for the individual venturis averages +/-0.72%. Again this figure excludes the uncertainty of the calibration facility (4-0.15%) which is not pertinent to the comparison of the UFM and the venturi data. It also does not include explicit allowances for calibration biases due to changes in throat diameter or tap geometry caused by the deposition of corrosion products. Neither does it include an explicit allowance for separation effects that may occur at Reynolds Numbers above those at which the modified venturi is calibrated. The uncertainty quoted does include allowances the secondary standard used to measure differential pressure during calibration, the modeling uncertainty as evidenced by the spread in data for various flow feed fractions in the test model, the uncertainty of the data fit, the random uncertainty in the Reynolds Number Extrapolation, and the systematic uncertainty in Reynolds Number Extrapolation (the latter is a measure of the spread in the two approaches to the extrapolation process). For the comparison with the UFM, this last uncertainty component will be removed, since one of the purposes of the comparison is to establish an appropriate extrapolation basis for these devices. When the systematic uncertainty in Reynolds Number extrapolation is removed, the differential uncertainty of the modified venturi discharge coefficient is diminished to +/-0.24%. If then the UFMs are used to calibrate the modified venturis at power the aggregate uncertainty of a calibration point is the root sum square of the reduced venturi differential uncertainty (+/-0.24%), the UFM differential uncertainty(+/-0. 14%), and an allowance for the transmitters used to read out the differential pressure in the plant. A figure of 4-0.25% will be used for this last element. A larger figure is appropriate for long term use, to account for drift and other effects, but the transmitters were calibrated just prior to plant startup and the 0.25%

figure is considered reasonable for this purpose. The aggregate uncertainty in each calibration point is thus [(0.24)2 + (0.14)2 + (0.25)2] = +0.37%.

TP77 Rev. 1 (13)

Figures 5A through 5D show what the respective UFM measurements in the plant indicate the discharge coefficients should have been; these data are the red squares shown on each figure. Despite the uncertainty in these points, the data of Figures 5A, 5B and 5C indicate that a log linear fit is a far better extrapolation basis that the reciprocal square root fit. This conclusion is supported not only by data at full power (Reynolds Number II million) but also, for Loops A and B, at 50% power (Reynolds Number 4 million). [No data were obtained for loop C at reduced power.] It should be pointed out that the reciprocal square root fit, which is somewhat more commonly used for extrapolation is non conservative with respect to the determination of power. The data of Figures 5A, 5B and 5C indicate the non conservatism is in the order of 0.5%

The data for the Loop D modified venturi, Figure 5 D, differ significantly from the other three flow tubes. Here a significant, non conservative bias is present-roughly 1% above the log linear fit at full power. There is nothing in the laboratory calibration data for this flow tube that would suggest the imminent departure from the log linear fit. The difference cannot be explained by the differential accuracy of the UFM or the modified venturi. The deposition of corrosion products is not a likely cause since normally this phenomenon causes a shift in the Other direction, and no shifts are seen in the other three flow tubes, which are exposed to the same feedwater chemistry. What the data suggest (but do not prove) is that a separation "bubble" abruptly occurs in the D modified venturi at a Reynolds Number in the 3 to 4 million range which causes a 1% shift in the CD characteristic. As noted previously, such shifts have been seen elsewhere in similar flow measuring devices.

Figure 5A

,Loop A UVT Calibration.r.Tapset Average I $ f i ll I

I 1

1 1 1 I

I I

I1 1

I 111 I I11 I

I 1 11 1 I 11 1 I

1 11 1 1 I1 I 11 1 0.9950I I

11 1 I

I 1 1 I

1 1

I 1

1 1 1 1 1

I 1 1 0.9700 I

1 1 1 I

I I II II l1 I 1 1

I 1 11 I

I III I

I11 0.9 I 1,11 I

I I

I III I

I 111 11 I

I I

II*

I 1

11 0.9650 1

oI I

I tI I

1 1 1 I

I II I

11111I UI I

I I11" I

IIII 11 1 II I

I I

I II I 111 II II IIII I

l II I,

o

11111..

Ii..

1.

I II I H

I I II I

I I I I

I I

I I II III

~

I I

I 111WI I

1. I I

I I I

IX I

II III I

III*

I I

I I

11 1 I

I i

I I

r i I

I I

11 11 I

.1 I

11111II I.6 0

I 11i1 1

.1i I

I lfi ii! I.

I I

I I I I!

25-25-50 0 0.0-100 100-0-0 X I;SQRT(RN)

LOG-UNEAR (RN)

-- &- Implemented Cd 0

Carrected Cd, LEFM standard 0.1000 1.0000 10.0000 Reynolds Numbertl,000.000 100.0000 TP77 Rev. I (14)

Figure 5B Loop B UVT Calibration. Tapset average 1.0000 0.9950 II

]

I I

SI II I Il Ill

~9l Ilgil 3.6 I Ill I Ill Ill I I II I Ill II III' II liit I

a$li II 0100:l~

3 333 333 1133 3 I Ill c

0 OiJI I

I 3I I

3 I

I I3 I

]'- I I33 I

I I

1t I

I I

I 6

i t, I

I I I I

I I

33I333 I3 I

I I

II I

I 1

1 3

3 33I33 I

11 1

3 I11I1 3

33 II I

I I

0.9750, t

I I

I II I

I I

3.

II11 I:

11 11 1

I.I I I

I 3

1 1 I13 I

I I

lI I

I I

I I 1 1 I

1*11 I

1 1I I I I I

I I 1 I1I I33II333II I

II I

I I

I

1. I

.3 I

1 1

1 II 3

11 13 1

3 13 3

3 I

11 I

3 11 11 1

I II11 3

3 I

I I

I I

3 1

1 I I

3 1

3 I

I I 3 II I

III1 1

1 1

n "tKlI I

I 3 11 3

11 13311 I

3 1

1 6

100-0-0

• 1 0-0-100 33-0-07 25-25&50 x llsq1(RN) a log llnear (RN)

-- W-Implmnwtdad Cd 0

Cornacted Cd, LEFM stanldaid 0.1000 1.0000 10.0=00 Reynolds Numnberll.000,000 100.0000 Figure 5C Loop C UVT Calibration Data Tapset average E

C.)

333333 I

I 11III11 3

I 1 11111 I

IIl 0

.3 3

3 3 3 33I 3

I 3

1 1 1 1 13 1 I

3 3 3 I3 I3 I' 0.9850 3I 33 33 3

II 333 I3 3

II3 3

0.9750 I

31311I3l I7I 0.9700 L II I

0-0-100 50-0-50 Z

25-25-50 Log-Linear 1J.*qr(RN)

Implemented Cd LI Corroded, LEFM standard 0.1 I

10 100 Reynolds Numr dXl,0.00 TP77 Rev. I (15)

Figure 5D Loop 0 LIVT Calibration. Tapsat Average 1.0100 I

I I

I li"ii I

I I

I ll I

I I

I II I I

I 11 11 I

a al I.;

I a

a I II, 1.0050I I

I I

III I

11111 I

I I IIIII*

II 41.0*000I II I

I i

1111 I

aIa u

I I

II I

I I

I I

I I

I I

I

-0 50950 o.9900 1X

.ISqat.RN)

aa

• Log.4Irear(RN.)

Iii

-- I mplementedCd 0.9I I

I I

• I i

IF a

I I I III:

I I

I I I:

II I

I I

II:II 0":"

I l

I l

III; I

I.

1111 I

a I:

I II I:*5" 0.9750 I..

0.I5 11111 I

I I L ahl a

1ore te11,LF anln I.70 I

I. l III I

I I

I?-

I I

I II 0.1000 1.0000 10.0000 100.0000 Reynolds Numnber/l,000.000 Comparison of Other Venturi Tubes and Flow Nozzles with Chordal UFMs Over the years Caldon has collected data comparing the indications of the venturi tubes and flow nozzles used for the measurement of feedwater flow in nuclear power plants against the measurements of 4-and 8-path chordal meters. Such comparisons are not generally as accurate as the comparisons of this paper; they nevertheless provide a statistical insight into the potential uncertainties of nozzles and venturis in service. Sixty two such comparisons have been made. The results are plotted in Figure 6. The figure shows an approximately normal distribution whose mean is 0.08% above zero (nozzles greater than UFMs), with 2 standard deviations about the mean of about +/-1l.4%. This figure characterizes the root' sum square of the aggregate venturilflow nozzle uncertainties and the chordal UFM uncertainties.

The chordal meters are a mix of 4 and 8-path meters. The 4 path meters have mass flow uncertainties in the range of +-0.5%; the 8-path meters have uncertainties in the +/-0.3%

range. These figures include the uncertainty of the calibration facility, as is appropriate since the UFMs were not calibrated at the same time as the flow nozzles and venturi tubes. The aggregate uncertainty of the chordal ultrasonic flowrneters in the figure is estimated at about +/-0.4% (2 standard deviations). If the 0.4 % figure is used for the 1FM uncertainty, the distribution of Figure 6 implies a typical nozzle or venturi uncertainty of

[j42 2]114-0.4 ]"

1.3%. The uncertainty of the flow nozzle/venturi tube indication is made up of two principal components: (1) the differential pressure transmitter and the associated signal processing loop and (2) the discharge coefficient. If an uncertainty of, say, +0.8% is assigned to the transmitter and signal processing (most of the installations TP77 Rev. I (16)

of the figure did not employ digital signal processing), the =2.3% residual uncertainty for the nozzle/ measurements implies an uncertainty for the discharge coefficient of slightly greater than +/-1% (2 standard deviations). This result is entirely consistent with the performance of the modified venturis described in this paper.

Figure 6 Distribution of Differences In Venturls and Flow Nozzles versus Chordal Ultrasonic Flowmeters 62 samples 0.500 0.700 0.600 0.300 0.20 0.100 2.5 U Plant Data 10 Normel Olabribuli n centered at mm I 2

1.5 1

0.5

% Difference, Venturi 0

-0.5

.1

-1i5 or Nozzle minus Chordal UFM

.2

-2.5

3. CONCLUSIONS:

1. For the modified venturi tubes analyzed in this paper, a log linear fit of the laboratory calibration data is superior to a reciprocal square root fit in predicting the discharge coefficient at plant Reynolds Numbers. The reciprocal square root fit leads to a systematic non conservative bias in the discharge coefficient of about Y2%.

2. The data of this paper, as well as comparative data between chordal UFMs and nozzles and venturis in other nuclear installations support an allowance of at least 1% for the uncertainty in extrapolating discharge coefficient of nozzles and venturis from laboratory to nuclear feedwater conditions. This allowance is over and above any allowance for differences between lab and field hydraulics, which were explicitly modeled for the venturis analyzed in this paper.

TP77 Rev. 1 (17)

Acknowledgement The authors wish to acknowledge the contribution of Richard W. Miller, P. E., who reviewed and commented on this paper.

References (1)

International Vocabulary of Basic and General Terms in Metrology (VIM),

International Organization for Standardization (ISO)

(2)

Don Augenstein, Herb Estrada, Ernie Hauser, Caldon Inc, "Traceability of Thermal Power Measurements, Part 1, Chordal Ultrasonic Flow Measurements", American Nuclear Society, September, 2004 (3)

D. Halmi, "Metering Performance Investigation and Substantiation of the

'Universal Tube' (U.V.T.):

Part 1-Hydraulic Shape and Discharge Coefficient", Transactions of the American Society of Mechanical Engineers, Journal of Fluids Engineering, 1974 (4)

Fluid Meters, Their Theory and Application, American Society of Mechanical Engineers (5)

ASME Steam Tables, 1967 (6)

D. Halni, "Metering Performance Investigation and Substantiation of the

'Universal Tube' (U.V.T.) Part 2-Installation Effect, Compressible flow and Head Loss", Transactions of the American Society of Mechanical Engineers, Journal of Fluids Engineering, 1974 (7)

A. Ferron, "Velocity Profile Effects on the Discharge Coefficient of Pressure Differential Meters", Transactions of the ASME, September 1963 (8)

R1P. Benedict, "Generalized Fluid Meter Discharge Coefficient Based Solely on Boundary Layer Parameters", Transactions of the ASME, October, 1979 (9)

R. W. Miller "An Evaluation of the 16 inch Nozzle Sections Used for Feedwater Flow Measurement at the Harris Plant" (10)

H. Estrada, "A Limited Evaluation of Feedwater flow Element Calibration Shifts in Westinghouse Pressurized Water Reactor Plants", June 1978 (11)

H. Schlichting, Boundary Layer Theory, McGraw Hill TP77 Rev. 1 (18)

110 International Conference on Nuclear Engineering Tokyo, JAPAN, April 20-23,2003 ICONEl1-38533 EFFECTS OF VARYING HYDRAULICS ON THE CALIBRATION OF EIGHT PATH CHORDAL ULTRASONIC METERS E. Hauser H. Estrada Caldon, Inc.

Caldon, Inc.

1070 Banksville Avenue 1070 Banksville Avenue Pittsburgh, PA 15216 Pittsburgh, PA 15216 Phone: 412-341-9920, Fax: 412-341-9951 Phone: 412-341-9920, Fax: 412-341-9951 e-mail: ehauserc)caldon.net e-mail: hestrada*,caldon. net J. Regan Key Technologies, Inc.

40 East Cross Street Baltimore, MD 21230 Phone: 410-385-0200, Fax: 410-385-1114 e-mail: iregan(,kevtechinc.com Keywords: flow measurement, ultrasonic, measurement uncertainty, calibration.

INTRODUCTION Eight path transit time ultrasonic meters are being used in the US, Europe and Japan to support measurement uncertainty uprates of nuclear power plants. Four path meters are also being used for more limited uprates; the focus of this paper is on the performance of the eight path meters.

The power uprates rely on the demonstration of improved power accuracy to justify a reduction in the traditional 2% margin between operating power and the power at which loss of coolant accidents and other transients have been analyzed. The flow, density and enthalpy of the feedwater are key elements in the power calculation, and the eight path ultrasonic meters measure the flow and temperature from which these elements are derived.

Caldon's uncertainty analyses for uprates that employ I

these meters are rigorous carrying multiple elements in several categories (e.g., time measurements, length measurements, hydraulics). A key element in the uncertainty analysis is an allowance for the uncertainties that the feedwater flow profiles introduce in the meters' flow calibrations.

To minimize and bound this uncertainty calibration tests are performed on each eight path element to be used in an uprate application.

Calibration coefficients are defined as the ratio of flow indication by the calibration facility to the flow indication by the ultrasonic meter.

The calibration coefficients of eight path flow meters differ from one another, regardless of installation geometry, over a range of about 1%. Power uprate applications that use eight path meters require an accuracy in the range of +/-

03%. To obtain this accuracy, it is necessary to calibrate eight path flow meters against a traceable standard of Copyright C 2003 by JSME

highi accuracy. Accordingly, Caldon calibrates their flow elements at a certified laboratory. Calibrations are performed in straight pipe, and, in addition, in full scale models of the plant piping configuration in which they will be used, to establish the value of any bias that the flow profile specific to the application may introduce'.

For a power uprate, it is also necessary and appropriate to establish bounds for, and to limit the uncertainty introduced by plausible flow profile effects that may not be present in the laboratory model test. Such effects can arise because of the physical limitations of the laboratory or because of unforeseen perturbations in plant hydraulics versus those of the lab. During calibration tests, therefore, the model configuration is varied parametrically, to establish the bounds for and to limit the uncertainty that hydraulic variability may introduce in a calibration.

This paper describes extensive testing of a prototype eight path meter, results of which have be used to define the sensitivity of 9 path meters to broad variations in flow profiles, both axial and transverse and to establish a methodology whereby the impact of these changes on the uncertainty of the meters cans be minimized.

The test data include axial profiles varying from the rounded characteristic of developed flow in rough pipe to the nearly flat characteristic downstream of non-planar bends. Swirl, a globally rotating transverse flow pattern, ranges from near zero to 400% of the axial velocity in one configuration, the latter. The prototype test results have been compared with hydraulic variations that have been measured by production eight path meters -in a wide range of specific nuclear applications, to establish the bounding nature of the prototype testing.

CHARACTERIZATION OF FLOW PROFILES USING EIGHT PATH METERS Caldon's eight path ultrasonic meters are arranged 'in two planes of four chords each, at right angles to each other and at a nominal 450 with respect to the axis of the flow element. The eight path meter prototype is shown in Figure 1. Because orthogonal paths are paired in four chordal planes, transverse velocities projected onto each path pair offset, when the velocity measurements of a pair of paths are averaged. Hence the path arrangement makes the 8 path flow meter insensitive to variations in transverse velocity.

Figure 1. Caldon's Eight Path Meter Prototype The chordal arrangement of the paired paths provides axial velocity measurements for each chordal location.

As will be seen these data can be used to characterize

'the axial velocity profile. Transverse fluid velocities in the field of the measurement can also be established, using the differences in the fluid velocities measured in each chordal plane.

The chordal arrangement of Caldon's eight path ultrasonic flow meter permits the shape of the axial velocity profile to be characterized using the ratio of the average of the velocities measured along the outside (short) chords to the average of the velocities measured along the inside chords. This ratio, called the flatness, can be used to predict the performance of ultrasonic meters in both eight path and single diametral path configurations. This flatness ratio defines how flat a flow profile is as compared to other measured profiles.

The flatter the velocity profile, the higher the flatness ratio. A perfectly flat profile has a flatness of 1.0.

Developed turbulent flow profiles in straight pipe with high relative roughness or low Reynolds number will have a flatness in the 0.75 to 0.8 range. Developed profiles at high Reynolds number in smooth pipe can produce a flatness of up to 0.9.. Downstream of nonplanair bends and similar features, flatness can approach 0.95. For the feedwater flow measurements associated with power uprates, the flatness for actual profiles measured in service have ranged from 0.81 to 0.95.

As has been noted, an eight path meter can also be used to quantify the transverse velocities present at a specific hydraulic location.

Swirl, a

globally rotating.

transverse flow pattern depicted schematically in Figure 2, is measured with an eight path meter using one half 2

2 Copyright 0 2003 by JSME

of the difference in the velocities measured along the outside chords. It may be shown that the mean tangential velocity at the location of the outside chords is equal to this difference.

As with flatness, swirl can affect the calibration of flow measurement systems (e.g.,

flow nozzles).

=rkpoftnt Figure 2.

Depiction of Swirl In Pipe Flow CALIBRATION DATA FOR THE PROTOTYPE METER Extensive tests of the Caldon eight path prototype meter show that the calibration of this flow element is not very sensitive to the shape of the axial profile. Figure 3 shows a linear best fit of calibration coefficients for this meter versus flatness of the flow profile. These data were obtained in a broad range of hydraulic configurations. The low flatness data were obtained in straight pipe with a variety of upstream flow conditions at varying distances from the flow element. The high flatness data were obtained in hydraulic configurations dominated by inertial forces, at varying distances downstream of single and compound bends, planar and non planar.

Over the extreme flatness range of the tests (flatness ratios of 0.81 to 0.95), the nominal change in the prototype meter calibration is less than 0.05%. A difference in calibration test flatness versus a measured flatness after installation in the plant of 0.04-relatively large based on present experience-would, according to the best fit of the data produce a calibration bias of less than 0.02%. The downward trend of the calibration coefficient with increasing flatness is generally in accordance with theory, reference (1). The slope in this case is somewhat lower than that calculated in the reference.

Likewise the eight path calibration is insensitive to swirl.

Figure 4 plots calibration factor against swirl, as measured by the tangential velocity (normalized to the average axial velocity) at the outside paths. The linear fit of the data indicates that an extreme swirl of 40%

produces only a 0.18% reduction in meter factor. This reduction is almost certainly due to the flatness of the profile in the maximum swirl configuration. The slope of the linear fit is negative because of the increasing flatness that generally accompanies swirling flow.

Suppose a difference of 10% in the swirl present at calibration versus the swirl present at the installed location in the plant. With the prototype 8 path meter, the resultant bias in-plant would be less than 0.05%.

Additionally, the differences in the profiles can be characterized by their flatness and, as will be discussed in the next section, the measurement of flatness in situ allows a small correction to be made to the calibration coefficient.

LAU,

kil I loll '

Lin,

Biil,

Bil A j

A A

AA uA

4iA, A

s XAAK gin fill Ism IMg ur N

I3. U Cli io Coefficients IVerI Fln esi Figure 3.

Calibration Coefficients Versus Flatness IJill A

A IR i

Ji~ll JIllll It It LIt Il

13 Iit lit Isi Ail ask Figure 4. Calibration Coefficients Versus Swid 3

Copyright 0 2003 by JSME

The insensitivity of the 8 path prototype flow element to velocity profile is confirmed by calibration experience for flow elements other than the prototype. Calibration data for Caldon CheckPlus flow elements produced for nuclear customers show that the calibration factors for a meter are generally insensitive to the profile flatness and that the flatness range measured for parametric variations in hydraulic configuration in the calibration lab bounds any difference between the nominal plant configuration in the lab and the plant. Furthermore, as will be discussed in the next section calibration test data, in combination with in-plant measurements can be used to limit uncertainties due to differences between calibration and plant profiles still further.

USE OF PLANT DATA TO ENHANCE THE ACCURACY OF THE CALIBRATION COEFFICIENT When a Caldon eight path meter is installed in a nuclear plant feedwater system, the in situ flatness is calculated from the individual path velocities, and compared to the flatness measured during calibration tests. The purpose of this measurement is to ensure that the calibration factor chosen based on the testing is appropriate for the installed conditions in the plant.

This process is illustrated using the data for an eight path LEFM CheckPlus system recently commissioned at a large Pressurized Water Reactor plant.

The eight path flow meter for this unit was calibrated in a model of the hydraulic configuration of the unit's feedwater system.

The feedwater model configuration was varied parametrically to provide reasonable assurance that the actual plant flow profiles would be bounded by the calibration data. The parametric test data showed only a small variation in calibration coefficient with flatness-over a range from 0.84 to 0.90, the variation in was +/- 0.1% about the mean. The mean calibration coefficient for all of the parametric data, 1.0022, was chosen as the value to be used in the plant.

Data from the plant, taken following commissioning, showed that the plant profile was flatter than the mean flatness for the parametric tests. Specifically the flatness in the plant is 0.90, at the upper end of the range for the calibration tests. The average flatness of the five calibration tests used to determine the profile factor was 0.88. The increased flatness in the plant is almost certainly due to increased swirl, which tends to flatten the velocity profile. The swirl in the plant, as measured by the tangential velocity at the 0.86 of the interior radius is 5% of the mean axial velocity. The mean swirl for the calibration tests was 2.61/a, with a range from 0 to 5% depending on the piping configuration tested.

As has been noted, the theoretical relationship between the shape of an axial velocity profile and the flow measurement of a 4 or 8 path chordal meter, reference (1), predicts that, as the velocity profile becomes flatter, the profile (meter) factor becomes slightly lower. This trend was seen in the prototype flow element discussed in the preceding section. Calibration data for this PWR flow element, including data for tests in straight pipe and for parametric variations of the plant model configuration, confirm the trend of reducing profile factor with increasing flatness. These data are plotted in Figure 5.

I=

I m

tu Lu

a.

L1 on M

I Figure 5.

Large PWR, Calibration Coefficient Versus Flatness Since the actual flatness of the profile at the plant was established by measurement in-plant, the least squares fit of the calibration data could be used to establish the calibration coefficient for the flatness in the plant. This point is plotted on Figure 5 (the "Plant datum"). Based on the fit and the measured flatness, the profile factor appropriate to this PWR is 1.0015-about 0.07% less than the factor implemented at the time of installation and commissioning. It should be noted that the calibration coefficients for the two calibration test configurations that produced a flatness comparable to the plant's (about 0.90) are also close to this figure.

However, the use of the fit is considered more appropriate, since the fit utilizes all of the calibration data. Accordingly, the calibration coefficient input to the LEFM CheckPlus at the plant was revised to 1.0015.

4 Copyright 0 2003 by JSME

The uncertainty in this profile factor (that is, the uncertainty in the fit of the data) is +/- 0.04% (2a). The use of the calibration coefficient based on the in situ flatness does not increase the uncertainty in the calibration coefficient and may, in fact, decrease it. This follows because the uncertainty in the mean of the parametric tests (which was used as a basis for the commissioning calibration) is essentially the same as the uncertainty of the least square fit used to determine the calibration coefficient appropriate to the in situ flatness.

Furthermore, the bounding analysis used to support the uprate carries an allowance for the uncertainty in the calibration due to the extrapolation from calibration laboratory conditions to plant conditions (the Reynolds number in the plant may be a factor of 6 to 10 greater than that in the lab). But the use of the in situ profile to correct the calibration coefficient arguably involves no extrapolation.

CONCLUSIONS

1. Extensive testing of an 8 path prototype meter in a broad spectrum of hydraulic configurations confirms the general insensitivity of 8 path chordal systems to axial and transverse fluid velocity profiles.

2.

An 8 path chordal system provides a

quantitative measurement of the axial profile, specifically, the flatness-the ratio of the axial velocities measured by the short (outside) chords to the axial velocities measured by the long (inside) chords. This information allows a quantitative assessment of the differences in hydraulic profile seen by a meter in a plant versus the hydraulic profile seen by that same meter in the calibration lab.

3.

Data for a prototype 8 path meter and for a typical S path meter now installed in a large PWR show small downward trends in calibration coefficient with increasing flatness.

The trends are generally in accordance with theory, reference (1).

4.

The sensitivity of an 8 path meter's calibration to flatness can be established quantitatively by parametric variations of the hydraulic configuration during calibration testing.

When this meter is installed in the field, the flatness measured in the field can be used with the calibration coefficient versus flatness relationship established in the lab to determine a calibration coefficient precisely adapted to the field application.

REFERENCES

1. Caldon Engineering Report, ER 262, Effects of Velocity Profile Changes Measured In Plant on Feedwater Flow Measurement Systems, January 2002 (Available on the Caldon

Website, www.Caldon.net) 5 Copyright 0 2003 by JSME

THEORY OF ULTRASONIC FLOW MEASUREMENT-GASES AND LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

Introduction Ultrasonic flow measurement systems (UFMs) are being applied with increasing frequency to hydrocarbon flow measurements. Most of these UFM s are transit time (also called time-of-flight) systems-they measure the transit time of ultrasonic energy pulses traveling with and against the direction of flow. This paper will outline the principles of three kinds of transit time UFMs:

Externally mounted ("strap on") transit time meters measuring liquid flow. In meters of this kind, the ultrasonic pulses travel through the liquid on a path at an angle determined by the physical properties of the liquid, the pipe on which transducer assemblies are mounted, and the mounting hardware.

"Chordal" transit time meters measuring liquid flow.

In meters of this kind, the transducers are installed in wells, similar to the thermrowells that are sometimes used to house RTDs or thermocouples.

The angles of the acoustic paths in these meters are determined by the mechanical design of the transducer wells and the spool piece in which the wells are mounted. The term "chordal* is used here because, in Caldon's designs of meters of this type, the acoustic paths are arranged in parallel chords across the spool. Other manufacturers arrange paths differently, but unless otherwise noted, the discussion will generally apply to their meters as well.

Chordal meters measuring gas flow. Mechanically, these meters resemble chordal meters that measure liquid flow. But different factors affect the performance of UFMs for gas, and they merit separate discussion.

It will be noted that there will be no coverage of externally mounted UFMs measuring gas flow. The technological challenges confronting the design of such meters are formidable (as will be evident from the discussion that follows). A few manufacturers provide external meters for a limited range of gas applications, but they have not found wide use.

Discussion Transit 77me Measurement Fundamentals A transit time ultrasonic flow measurement system transmits acoustic energy along one or more diagonal paths through the pipe in which flow is to be measured.

Such an acoustic path is illustrated in Figure 1. In the configuration shown, a pair of transducers are mounted to form a diametral diagonal path through flowing liquid, but the fundamental principles described in the following paragraphs apply to gas and liquid, internal or external.

If the upstream (A) transducer is excited by a burst of electrical energy, it will transmit a packet or pulse of mechanical (acoustic) energy into the adjacent medium. In Caldon's LEFMs, the electrical excitation of the transducer also initiates a time measurement by causing counts from a precision electronic clock to be accumulated in a counter. The pulse of ultrasound will consist of several cycles having a frequency typically in the 0.5 to 3 megahertz range for liquid flows, and in the 50 to 500kilohertz range for gas flows. The transducer is usually designed to be directional, so, in the configuration illustrated in the figure, a significant fraction of the acoustic energy will travel in a straight line from transducer A to transducer B, where it will produce a small burst of electrical energy. If the arrival of the energy at transducer B is detected with suitable electronics and this detection causes the accumulation of clock pulses in the time counter to stop, the elapsed time tA, from the time of transmission to the time of detection, has been measured (by the number of clock pulses accumulated).

If, now, the downstream or B transducer is excited and the arrival of acoustic energy at transducer A is detected, the transit time taA can be measured in like manner. The measured times are related to the dimensions, properties and velocity of the fluid as follows:

1) t,*= [Lptm I (cpm + vpth)] + *.a fld dely

2) t8A, = [.

I (cpt - vpth)] + I..

fluid delay.

Where Lpo is the length of the acoustic path, cp~.t Is the mean ultrasound propagation velocity along the acoustic path with the fluid at

rest, Vpath is the mean fluid velocity projected onto the acoustic path, and Tnon fluid deaiy is the total of the electronic and acoustic delays exterior to the fluid.

Each energy pulse traverses exactly the same path in the non fluid media and, in Caldon's LEFMs, the same transmitter produces each pulse and the same electronic detector detects each pulse. Consequently, the difference in the transit times, At, is given by:

3A) At = tBA - tAB

= [ Lp., I ( cpat - v,.th) ] " [

/L I (c., + v,.t) ]

Putting both terms over a common denominator and performing the algebra:

3B) At = 2 Lpth vp~th / (Czh 2 _ Vpth2)

THEORY" OF ULTRASONIC FLOW MEASUREMENT-GASES AND L:IQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

In most liquids the sound velocity is two orders of magnitude larger than the fluid velocity, c ranging from 2500 ft/sec to 5500 ft/sec versus v of 2 to 30 ft/sec.

Hence equation 3B can be approximated 3C) At = 2 Lp=t vp.h / cpah 2-Or 3D) vpt -=_

At c=,2 / (2 Lpo1)

Some eady UFMs had the user input sound velocity from a look-up table in equation (3D) to find path velocity. This procedure is not consistent with good accuracy.

In most liquids, sound velocity varies strongly with temperature and weakly with pressure.

Hence varying liquid product temperature renders the meter calibration invalid.

If sound velocity is determined from transit time by. one of the methods described in a later paragraph, equation (3D) is an acceptable approximation to determine path velocity in liquids. Even with a relatively compressible hydrocarbon like liquefied natural gas (with, therefore, a relatively low sound velocity) the error due to v2 is unlikely to exceed 0.01%. However, the approximation of equations (3C) and (3D) is usually unacceptable for gas flow. Here neglecting the v2 term

-can introduce velocity-dependent errors of 1% or more.

For precision, therefore, a gas UFM must use its transit time measurements to determine (c2 - v2) as well as A.L The transit times in the fluid, tgAa and tBA are found by subtracting the non fluid delay from the measured transit times. For a given application, the non fluid delay Tnon Wdfdeiay may be calculated or measured (or both).

4A) ttB t4B - "*non 11uid d.ely 4B) tMA tBA - tnon fluid delay The product of these fluid transit times yields the following:

5) tfAB tA = [Lp.1h / (cpth + vpam)] x [L, I c(pth - Vpt)]

= L.,. 2I (c*, 2* - v., 2=)

Combining equations 3B and 5, the following expression is obtained for the product of the acoustic path length and the fluid velocity projected onto the path.

6)

Lpth Vp6th = (%)1-m.2 At/( tIAS t1BA)

This relationship is fundamental to the operation of all transit time flowmeters. Essentially it says that the product of the path length and the mean velocity along that path can be determined by transit time measurements with an absolute accuracy limited only by

" The accuracy of the transit time measurements

" The accuracy of the measurement (or calculation) of the non fluid time delay The accuracy of the path length measurement This is of course only a statement about the accuracy of a path velocity measurement-not volumetric flow.

The accuracy with which one or more of these path velocity measurements gets translated into volumetric flow is affected by other factors, both acoustic and hydraulic. These factors will be covered in later discussion.

Note that the sound velocity can also be determined from the measurements of the transit times by substituting the fluid transit times in equation 5. [The v

-term can be calculated using equation 6 or, if it is small compared to c2, neglected.] The sound velocity of a product is a state variable like temperature and pressure and in a pipeline carrying a single product can be used with pressure to determine temperature.

Alternatively,. in multiproduct pipelines, sound velocity can be used

alone, or with a

temperature measurement, to detect product interfaces.

Figure I Geometry of a Transit Time Acoustic Path'

'MONS~uV

-L PA7H PI EN1~HAL-L4TRANSfOJr!R B 0- PATH ANGLE How accurately can the fluid velocity projected along the acoustic path be measured using equation (6)?

Essentially, with an accuracy determined entirely by the accuracy of the measurements of the transit times and the separation distance, and the accuracy of the measurement or calculation of the non fluid delays.

THEORY OF ULTRASONIC FLOW MEASUREMENT--GASES AND LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

Some Numbers How big are the times and time differences that UFMs measure? Suppose a 2-path chordal UFM with a path angle 0 of 450 is measuring crude oil flow in a 12 inch pipeline. Petroleum product sound velocities usually lie in the range of 2700 ft/sec to 5000 ft/sec. If a sound velocity of 4500 ft/sec is assumed (typical of a medium crude), the transit times will be about 280 psec. The time difference, At, at rated flow will equal 430 nanoseconds (1 nanosecond = 10"Q seconds), for a pipeline velocity of 5 ft/sec. If a 10:1 turndown is specified for this meter, the At at the low end of the flow range will be 43 nanoseconds.

The transit time of an external UFM, like that in Figure 2, may be slightly smaller than the chordal example because physical properties of the pipe and fluid dictate a shallower path angle. With.typical petroleum product properties and steel pipe, the angle will be about 20°. [How the path angle of an externally mounted UFM is determined will be described later.]

The transit times for an external meter mounted on the same 12 inch pipe will lie in the 250 lpsec range. The At at rated flow of 180 nanoseconds. [To increase the magnitude of the At many externally mounted UFMs are configured in a 'bounce" or V mode, wherein the two transducers are mounted on the same side of the pipe and the acoustic path length is doubled. This arrangement doubles both the t and At.]

Clearly, one of the challenges of a UFM measuring liquid flow is the accurate measurement of very small times and particularly time differences (At). For a 10:1 turndown and a linearity of 0.2%, the chordal UFM described above must measure time differences with an accuracy of +/- 90 plcoseconds (1 picosecond = 1 x 10.12 seconds). The externally mounted UFM must do even better--it must measure time differences with an accuracy of. +/- 35 picoseconds if it is configured in the direct mode (as in Figure 2 below) and +/- 70 picoseconds if it is configured in the bounce mode.

Some UFMs achieve these accuracies and better. To do so, their designers must pay particular attention to what is called the reciprocity of the signal processing that they use-the non fluid delays must be exactly the same in the upstream and downstream direction.

Signal quality is also essential-here, elimination of noise is the key.

There are different challenges for the designers of UFMs that measure gas flow. Here the transit times and At's are several orders larger than for meters measuring liquid flow. For example, the transit times for a two path chordal meter measuring the flow of natural gas in a 24 inch pipeline would be around 1.75 milliseconds. At rated flow, the time difference (At) would lie in the 100 to 200 psecond range, depending on pipeline velocity. A major challenge in gas flow measurement lies in reliably detecting a relatively small ultrasonic pulse, possibly in the presence of noise.

Dealing with wide variations in transit times due to turbulence and other factors is also more difficult in gas versus liquid meters.

The small size of received pulses in ultrasonic gas flow measurements is the inherent result of what is called the acoustic impedance mismatch between the transducers and the flowing medium. Because the pulse-producing transducer is relatively dense and stiff and the flowing medium is relatively light and compressible, most of the acoustic energy reaching an

• interface between the two stays where it started. That is, a large fraction of the energy is reflected rather than transmitted. There are at least two such interfaces in every acoustic path. Pulses traveling liquid paths also are attenuated at interfaces, but the degree of attenuation is several orders less challenging in the

-liquid case.

Translating Path Velocities into Axial Velocities and Volumetric Flow All of the preceding describes a methodology for measuring a fluid velocity projected onto an acoustic path. To determine volumetric flow rate from one or more sets of path measurements requires that (1) the path velocity (or velocities if more than one measurement is made) be related to the axial fluid velocity which produced it, and (2) the axial fluid velocity for, the acoustic path (or paths, if there Is more than one) be related to the mean axial velocity for the pipe cross section.

The first of these conditions requires a knowledge of the angle 4 between the acoustic path and the pipe axis, illustrated in Figure 1. It also requires a knowledge of the fluid velocity component normal to the pipe axis, If there is any (i.e., the transverse fluid velocity). The projection of the axial fluid velocity onto the acoustic path is shown in Figure I. No transverse velocity component is. shown in the figure; its impact will be discussed later. From the trigonometry:

5) vpat = vaxi sin.

Where vdw is the mean axial fluid velocity projected along the acoustic path, and 4 is the angle of the acoustic path through the fluid, measured from the normal to the pipe axis.

THEORY OF ULTRASONIC FLOW MEASUREMENT--GASES AND LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

Equation 4 can be rewritten in terms of the axial fluid velocity in the way of the acoustic path:

6A) v puh = Vaia sin

=At cpa/ 2 Lpt h) 6B)

Vaxil = At Cpat2 / (2 Lp.t sin 1 )

The specifics of how the path angle is determined and how one or more axial velocity measurements along the path(s) are translated into volumetric flow depends on whether the meter is external or chordal, and if chordal, the arrangement of the chords. The external meter will be covered first.

Principles of Externally Mounted Transit time Systems In an externally mounted

UFM, Snell's Law of Refraction constrains the geometry of the path traveled by acoustic pulses through the flowing fluid.. Essentially these pulses must travel in a diametral plane. Such a configuration is shown in Figure 2. Here the path length is related to the internal diameter of the pipe, ID, by

7)

L p~th = IDIcos4.

For this configuration, from equations (65) and (7), the axial velocity averaged over the. diametral acoustic path is given by

8)

.Va4a c

iaAt cpath/

2 IDtan )

This is the governing equation for externally mounted transit time ultrasonic flowmeters, in the absence of transverse flow. As has been noted, the acoustics of.

the pipe wall and fluid require placement of the transducers for such meters on diametral diagonals; hence, externally mounted ultrasonic flowmeters are essentially velocimeters. From the velocity measured in accordance with equation 8, the flow must be determined.

It should be pointed out that for externally mounted transit times ultrasonic systems, the path angle 0 is not simply determined by transducer placement. Figure 2 provides a picture of external system acoustics.

Piezoceramic transducer elements are mounted on wedges which, in turn, are mounted on the exterior of the pipe. The wedge optimizes the acoustic interfaces between the transducer-wedge assembly and the pipe wall and between the pipewall and the fluid. The three angles of the ray path in Figure 2, ýF, 1p and 4kw are the path angles followed by the pulses in the fluid, pipe, and wedge respectively. The angle OF is equivalent to 0, the angle through the fluid, that has been used in the discussion of Figure 1. The wedge, the pipe, and the fluid angles are all governed by Snell's law of refraction. They are also affected by the size, placement, and configuration of the wedges. Snell's law stipulates that

9) sin 4F/IcF = sin p/cp =sin 43/c.

Where cF, cp and c, are the respective sound velocities of fluid, pipe, and wedge.

If the three sound velocities are measured or otherwise determined, it remains only to establish one of the three angles. The angle ow would seem to be the obvious choice-to determine the others, and thus the acoustic path through the fluid, since the wedge can be manufactured with a precise geometry.

Figure 2 Acoustics of an Externally Mounted Transit time

• UFM 0F -LU0O PA1 A*G LE But determining the exact angle of the path in the fluid from the wedge angle is not always straightforward. If the transducers are acoustically distant from one another, ýw can be determined by assuming the path connects the centers of the piezoceramic elements (refer again to Figure 2). Note that in this case, the ray path Is not necessarily perpendicular to the transducer face; hence the wedge angle is not necessarily equal to the mechanical angle of the sloping face.

On the other hand, if the transducers are acoustically close to one another, 43, is determined by the mechanical configuration of the wedge; it is the angle between a normal to the transducer transmitting surface and a normal to the axis of the pipe. Often, the acoustics are such that neither assumption is exactly valid, and both wedge configuration and transducer placement affect the path angle through the fluid.

THEORY OF ULTRASONIC FLOW MEASUREMENT--GASES AND LIQUIDS Class 3190:

Herb Estrada, Chief Engineer, Caldon, Inc.

Returning to equation. B, it can be seen that the accuracy of the velocity measurement of an externally mounted transit time system is a function not only of the accuracy of the time, distance and non fluid delay measurements, but also of the accuracy with which their acoustics can be characterized. The answer one obtains from equation 8 is very sensitive to the tangent of the angle o, An accurate fluid sound velocity measurement is.

crucial to establishing the path angle C. To enhance the accuracy with which fluid sound velocity is determined in its external meters, Caldon employs a second pair of transducers, mounted so as to form an acoustic path normal to the pipe axis (the 'cross path" in Figure 2).

This arrangement is inherently less susceptible to variations in the physical properties and dimensions of the pipe than is the diagonal path. Data from this path can also be used to compensate for transverse flow, as noted below.

The variable of interest is volumetric flow--not velocity.

Volumetric flow. Q is given by IOA)

Q = (pipe cross sectional area) vman, a),,

where. v mean axial is the mean or average, fluid axial velocity over the internal pipe cross sectional area.

10B)

Q = [1 ID2/4]v vm,,,

For the determination of volumetric flow from an acoustic system with transducers on a diametral diagonal as they are in an externally mounted UFM, it thus remains to relate the diametral axial velocity to the axial velocity averaged, over the pipe cross section.

The two velocities are rarely the same. In a long straight section of feedwater pipe at Reynolds numbers in the 106 range, the velocity measured along a diametral diagonal will typically be greater than the true mean velocity by 5 or 6%. The exact number depends not only on kinematic viscosity, diameter and velocity (that is, the Reynolds Number) but also on relative roughness of the pipe wall. At a Reynolds number of 104, the measured velocity may be 10% or 12% greater than the true mean. In the laminar flow regime it is 33%

greater.

On the other hand, a short distance downstream of a header the measured velocity and mean velocity may be within 1 or 2% of each other.

Summing up, in a specific application, meter calibration may vary with:

product (because viscosity and hence Reynolds Number varies),

velocity (which is also an element of Reynolds Number),

pipe condition (because velocity profiles vary with relative roughness as well as with Reynolds Number), and with hydraulic configuration (because this too affects velocity profile).

The differences between diametral axial velocity and mean axial velocity arise because of the differences in the shapes of the Velocity profiles. The diametral diagonal paths of externally mounted ultrasonic meters undersample the region near the pipewall relative to its area, and oversample the region near the middle of the pipe relative to its area.

Caldon ultrasonic systems use a profile factor, PF, to relate the axial fluid velocity measured along one or more acoustic paths to mean axial fluid velocity.

Specifically I1A)

V,en..al=

(PF) V aa.path

Hence, 1.1B)

Q = [I IDD/4] (PF) At CF2/ (2 ID tan 4F )

Equation 11B is used by Caldon for.externally mounted systems operated in the direct mode, as in Figure 2. These meters can produce excellent linearity and repeatability, providing the range of Reynolds number coverage is not too broad.

As has been noted, the inference of axial velocity from diagonal path At (implicit in equation (11B)) is only valid in the absence of significant transverse velocity.,

Unfortunately, transverse velocity is sometimes present in locations where it is practical to install an externally mounted ultrasonic system. Caldon LEFMs deal with transverse velocity in one of two ways:

(1) The time differential from a path normal to the pipe axis (which path. is also used to determine fluid sound velocity) is used to calculate transverse velocity and the result is subtracted from or added to the path velocity as appropriate, or (2) The diagonal path is configured in the 'bounce' mode. That is, both diagonal path transducers are mounted on the same side of the pipe so as to form a V-shaped acoustic path through the fluid. In this configuration, the transverse velocity projection on one leg of the V (relative to the axial component) is offset by the approximately equal and opposite projection on the other leg. For this mode, the divisor of equation 11B is doubled (because the acoustic path in the fluid is twice as long).

THEORY OF ULTRASONIC FLOW MEASUREMENT-GASES AND LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

Some Numbers How big are the times and time differences that UFMs measure? Suppose a 2-path chordal UFM with a path angle ý of 450 is measuring crude. oil flow in a 12 inch pipeline. Petroleum product sound velocities usually lie in the range of 2700 ft/sec to 5000 ft/sec. If a sound

-velocity of 4500 ft/sec is assumed (typical of a medium crude), the transit times will be about 280 gisec. The time difference, At, at rated flow will equal 430 nanoseconds (1 nanosecond = 10"' seconds), for a pipeline velocity of 5 ft/sec. If a 10:1 turndown is specified for this meter, the At at the low end of the flow range will be 43 nanoseconds.

The transit time of an external UFM, like that in Figure 2, may be slightly smaller than the chordal example because physical properties of the pipe and fluid dictate a shallower path angle. With typical petroleum product properties and steel pipe, the angle will be about 20.

.[How the path angle of an externally mounted UFM is determined will be described later.]

The transit times for an external meter mounted on the same 12 inch pipe will lie in the 250 lisec range. The At at rated flow of 180 nanoseconds. [To increase the magnitude of the At many externally mounted UFMs are configured in a 'bounce" or V mode, wherein the two transducers are mounted on the same side of the pipe and the acoustic path length is doubled. This C

arrangement doubles both the t and At.]

Clearly, one of the challenges of a UFM measuring liquid flow is the accurate measurement of very small times and particularly time differences (At). For a 10:1 turndown and a linearity of 0.2%, the chordal *UFM described above must measure time differences with an accuracy of +/- 90 picoseconds (1 picosecond = 1 x 10-12 seconds). The externally mounted UFM must do even better-it must measure time differences with an accuracy of +/- 35 picoseconds if it is configured in the direct mode (as in Figure 2 below) and +/- 70 picoseconds if it is configured in the bounce mode.

Some UFMs achieve these accuracies and better. To do so, their designers must pay particular attention to what is called the reciprocity of the signal processing that they use-the non fluid delays must be exactly the same in the upstream and downstream direction.

Signal quality is also essential-here, elimination of noise is the key.

There are different challenges for the designers of UFMs that measure gas flow. Here the transit times and At's are several orders larger than for meters measuring liquid flow. For example, the transit times for a two path chordal meter measuring the flow of natural gas in a 24 inch pipeline would be around 1.75 milliseconds. At rated flow, the time difference (A) would lie in the 100 to 200 psecond range, depending on pipeline velocity. A major challenge in gas flow measurement lies in reliably detecting a relatively small ultrasonic pulse, possibly in the presence of noise.

Dealing with wide variations in transit times due to turbulence and other factors is also more difficult in gas versus liquid meters.

The small size of received pulses in ultrasonic gas flow measurements is -the inherent result of what is called the acoustic impedance mismatch between the transducers and the flowing medium. Because the pulse-producing transducer is relatively dense and stiff and the flowing medium is relatively light and compressible, most of the acoustic energy reaching an interface between the two stays where it started. That is, a large fraction of the energy is reflected rather than transmitted. There are at least two such interfaces in every acoustic path. Pulses traveling liquid paths also are attenuated at interfaces, but the degree of attenuation is several orders less chalFenging in the

-Jiquid case.

Translating Path Velocities into Axial Velocities and Volumetric Flow All of the preceding describes a methodology for measuring a fluid velocity projected onto an acoustic path. To determine volumetric flow rate from one or more sets of path measurements requires that (1) the path velocity (or velocities if more than one measurement is made) be related to the axial fluid velocity which produced it, and (2) the axial fluid velocity for the acoustic path (or paths, if there is more than one) be related to the mean axial velocity for the pipe cross section.

The first of these conditions requires a knowledge of the angle 0 between the acoustic path and the pipe axis, illustrated in Figure 1. It also requires a knowledge of the fluid velocity component normal to the pipe axis, If there is any (i.e., the transverse fluid velocity). The projection of the axial fluid velocity onto the acoustic path is shown in Figure I. No transverse velocity component is shown in the figure; its impact will be discussed later. From the trigonometry:

5) vp,* = vw., sin 0 Where v*,, is the mean axial fluid velocity projected along the acoustic path, and 4 is the angle of the acoustic path through the fluid, measured from the normal to the pipe axis.

THEORY OF ULTRASONIC FLOW MEASUREMENT-GASES AND LIQUIDS

.Class 3190 Herb Estada, Chief Engineer, Caldon, Inc.

(!-

Equation 4 can be rewritten in terms of the axial fluid velocity in the way of the acoustic path:

6A) v th = V",j sin =At cpth2 /2 Lp*)

6B)

V,.w3 = At Cpp 2 / (2 Lp=t sin )

The specifics of how the path angle is determined and how one or more axial velocity measurements along the path(s) are translated into volumetric flow depends on whether the meter is external or chordal, and if chordal, the arrangement of the chords. The external meter will be covered first.

Principles of Externally Mounted Transit time Systems In an externally mounted

UFM, Snell's Law of Refraction constrains the geometry of the path traveled by acoustic pulses through the flowing fluid. Essentially these pulses must travel in a diametral plane. Such a configuration is shown in Figure 2. Here the path length Is related to theinternal diameter of the pipe, ID, by

7)

L. P=,h = ID/ cos

• For this configuration, from equations (6B) and (7), the axial velocity averaged over the diametral acoustic path is given by

8)

V"=.a1At Cpýth(2 2 ID tan4*)

This Is the governing equation for externally mounted transit time ultrasonic flowmeters, In the absence of transverse flow. As has been noted, the acoustics of the pipe wall and fluid require placement of the transducers for such meters on diametral diagonals; hence, externally mounted ultrasonic flowmeters are essentially velocimeters. From the velocity measured in accordance with equation 8, the flow must be determined.

It should be pointed out that for externally mounted transit times ultrasonic systems, the path angle 0 is not simply determined by transducer placement. Figure 2 provides a picture of external system acoustics.

Piezoceramic transducer elements are mounted on wedges which, in turn, are mounted on the exterior of the pipe. The wedge optimizes the acoustic interfaces between the transducer-wedge assembly and the pipe wall and between the pipewall and the fluid. The three angles of the ray path in Figure 2, Or, Op and Ow are the path angles followed by the pulses in the fluid, pipe, and wedge respectively. The angle OF is equivalent to 0, the angle through the fluid, that has been used in the discussion of Figure 1. The wedge, the pipe, and the fluid angles are all governed by Snell's law of OF.FU PATH M4QGE Op-PIP ANGLE OW-EGE ANGL refraction.

They are also affected by the size, placement, and configuration of the wedges. Snell's law stipulates that

9) sin OF I CF = sin Op / cp = sin 0./r,,

Where cF, cp and c., are the respective sound velocities of fluid, pipe, and wedge.

If the three sound velocities are measured or otherwise determined, it remains only to establish one of the three angles. The angle Ow would seem to be the obvious choice-to determine the others, and thus the acoustic path through the fluid, since the wedge can be manufactured with a precise geometry.

Figure 2 Acoustics of an Externally Mounted Transit time UFM But determining the exact angle of.the path in the fluid from the wedge angle is not always straightforward. If the transducers are acoustically distant from one another, 4, can be determined by assuming the path connects the centers of the piezoceramic elements (refer again to.Figure 2). Note that in this case, the ray path is not necessarily perpendicular to the transducer face; hence the wedge angle is not necessarily, equal to the mechanical angle of the sloping face.

On the other hand, if the transducers are acoustically close to one another,

,, is determined by the mechanical configuration of the wedge; it is the angle between a normal to the transducer transmitting surface and a normal to the axis of the pipe. Often, the acoustics are such that neither assumption is exactly valid, and both wedge configuration and transducer placement affect the path angle through the fluid.

THEORY OF ULTRASONIC FLOW MEASUREMENT--GASES AND: LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

Returning to equation 8, it can be seen that the accuracy of the velocity measurement of an externally mounted transit time system Is a function not only of the accuracy of the time, distance and non fluid delay measurements, but also of the accuracy with which their acoustics can be characterized. The answer one obtains from equation 8 is very sensitive to the tangent of the angle Of.

An accurate fluid sound velocity measurement is crucial to establishing the path angle 0. To enhance the accuracy with which fluid sound velocity is determined in its external meters, Caldon employs a second pair of transducers, mounted so as to form an acoustic path normal to the pipe axis (the "cross path" in Figure 2).

This arrangement is inherently less susceptible to variations in the physical properties and dimensions of the pipe than is the diagonal path. Data from this path can also be used to compensate for transverse flow, as noted below.

The variable of interest is volumetric flow-not velocity.

Volumetric flow Q is given by 10,A)

Q = (pipe cross sectional area) Vm.an,axil where v ma,

- is the mean or average fluid axial velocity over the internal pipe cross sectional area.

S 101B)

Q-= [n ID /4] v mean axial For the determination of volumetric flow from an acoustic system with transducers on a diametral diagonal as they are in an externally mounted UFM, it thus remains to relate the diametral axial velocity to the axial velocity averaged over the pipe cross section.

The two velocities are rarely the same. In a long straight section of feedwater pipe at Reynolds numbers in the 106 range, the velocity measured along a diametral diagonal will typically be greater than the true mean velocity by 5 or 6%. The exact number depends not only on kinematic viscosity, diameter and velocity (that is, the Reynolds Number) but also on relative roughness of the pipe wall. At a Reynolds number of 104, the measured velocity may be 10% or 12% greater than the true mean. In the laminar flow regime it is 33%

greater.

On the other hand, a short distance downstream of a header the measured velocity and mean velocity may be within 1 or 2% of each other.

Summing up, in a specific application, meter calibration may vary with:

product (because viscosity and hence Reynolds Number varies),

velocity (which is also an element of Reynolds Number),

pipe condition (because velocity profiles vary with relative roughness as well as with Reynolds Number), and with hydraulic configuration (because this too affects velocity profile).

The differences between dlametral axial velocity and mean axial velocity arise because of the differences in the shapes of the velocity profiles. The diametral diagonal paths of externally mounted ultrasonic meters undersample the region near the pipewall relative to its area, and oversample the region near the middle of the pipe relative to its area.

Caldon ultrasonic systems use a profile factor, PF, to relate the axial fluid velocity measured along one or more acoustic paths to mean axial fluid velocity.

Specifically 11A)

V mean, axial = (PF) V 1ia.

path

Hence, 11B)

Q = [i ID 2I4] (PF)A t cF2/(2 ID tan O)

Equation 11B is used by Caldon for externally mounted systems operated in the direct mode, as in Figure 2. These meters can produce excellent linearity and repeatability, providing the range of Reynolds number coverage is not too broad.

As has been noted, the inference of axial velocity from diagonal path At (implicit in equation (111B)) is only Valid in the absence of significant transverse velocity.

Unfortunately, transverse velocity is sometimes present in locations where it is practical to install an externally mounted ultrasonic system. Caldon LEFMs deal with transverse velocity in one of two ways:

(1) The time differential from a path normal to the pipe axis (which path is also used to determine fluid sound velocity) is used to calculate transverse velocity and the result is subtracted from or added to the path velocity as appropriate, or (2) The diagonal path is configured in the 'bounce' mode. That is, both diagonal path transducers are mounted on the same side of the pipe so as to form a V-shaped acoustic path through the fluid. In this configuration, the transverse velocity projection on one leg of the V (relative to the axial component) is offset by the approximately equal and opposite projection on the other leg. For this mode, the divisor of equation 11B is doubled (because the acoustic path in the fluid is twice as long).

THEORY OF ULTRASONIC FLOW MEASUREMENT-GASES AND LIQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

To determine the profile factor PF of equation 11, the hydraulics at the location of the measurement must be characterized. Caldon draws on an extensive library of hydraulic model testing for external systems for this purpose. For readers interested in more detail on Caldon's experience in profile factor measurements for external

systems, Mazzola and Augenstein' is suggested.

Principles of Chordal (internal) Transit time Systems The discussion in the preceding section has focused on externally mounted LEFMs, where the acoustic paths are diametral and the acoustics themselves are determined by the properties and placement of transducer wedges and the dimensions and properties of pipe and fluid. It is now appropriate to consider the operative equations for a chordal or internal system. In these systems, transducers are inserted in wells that are, as noted before, somewhat similar to thermowells.

The ultrasound generated by a transducer passes through the "face" or *window* of the well in a direction normal to the face. Opposing transducer wells are located so that the centerlines normal to their faces coincide and form the nominal acoustic path.

This is the first of two important distinctions between external and chordal systems: the angle of the acoustic path In a chordal system is established mechanically by the angle formed by the centerline connecting the two transducer wells and the axis of the spool piece.

As a consequence, the path angle for a chordal system (or the angles for systems with multiple chords) can be established with an accuracy determined by dimensional control of the spool piece as opposed to the acoustics of wedges, pipe and fluid. Path angle is crucial to determining the axial velocity subtended by the acoustic path (as was shown in equation 6B). Since dimensions are typically controllable with much greater precision than acoustics, chordal systems possess an inherent accuracy advantage on this score.

In order directly to measure volumetric flow, one must integrate the axial fluid velocity over a cross section normal to the pipe axis, as illustrated in Figure 3. That is,

13)

Q V Wv (X, y) dx dy A four path chordal system approximates this double integration. To understand how, recall equation 6:

6)

Lp°h V 8

= (Y2)Lpth2 At /( tAB tfBA )

Also recall, from equation (5) 5)

Vpa = v,,I sin ý Refer now to the illustration of the four path chordal system in Figure 4. It will be seen that, for chord 1,

14)

L pm, = L d.,dj/cos ýj Substituting for vpth and Lp.th in equation (6), the following expression is obtained for chord 1:

15A) (v", 1 Lhord )(sin 0/cos ý) = (1/2)Lp 2th2 At /( t6 tisA) 15B) v=*= I Lcho,1 1 = (%)(Lpathv 21 tan 01)(At /(tfAE tfBA)

The LV product of equation 15B is exactly the line integral of V=,A dx at the location of chord 1. The chordal instrument illustrated in figure 4 performs four such integrations at locations yI,. y2, y3, and y4, effectively, dividing the pipe cross-section into four segments. The effective width of each segment is a fraction, w, of the internal diameter, ID, measured along the y axis.

Figure 3 Integration of Axial Velocity over a Pipe Cross Section I>

Treating the four chordal measurements as four elements of a numerical integration, the volumetric flow can be calculated as follows:

16) Q = ID [w, "w, vw, 1, + w2 Lchor 2 VaxW 2 +

W3 I-rOd vax,=,M + w4 Lccrd4 va,18l]

Or, substituting the lengths and times measured by the UFM in a more general expression:

17) Q = (ID12) ({

IN (wi Lpathl2/ tan 01)(Ati /(tfASI tfBAj)}

tt*.'F" I D.E. Mazzola and D.R. Augenstein, Hydraulic Testing of External Mount Ultrasonic Flow Meters, July 1995 where, in the four path system, the subscript can take on values from I through 4. Note that the times ttABI and tfsA1 in the above are transit times in the fluid;

THEORY OF ULTRASONIC FLOW MEASUREMENT-GASES AND UQUIDS Class 3190 Herb Estrada, Chief Engineer, Caldon, Inc.

non fluid delays must be determined and subtracted from the measured transit times to obtain the times used in this expression.

Figure 4 A 4 Path Chordal LEFM

(

-7 U

For Caldon chordal systems, the path locations, y, and weighting factors w are not chosen arbitrarily but comply with numerical integration rules specified by the mathematician Gauss2.

This integration technique will integrate polynomials up to the seventh order without error. Caldon has collected extensive calibration data for four path systems operating in a wide variety of hydraulic configurations. These data were obtained at a certified facility, for the most part at high Reynolds Numbers.

The data show that a meter factor in the 0.994 to 1.004 range is necessary, primarily to account for the difference between the circular geometry and the rectilinear geometry for which the Gauss procedure was developed. The data also demonstrate that the meter factor for four path Gaussian integration will handle a broad range of hydraulic geometries, with departures from nominal usually less than 0.2%.

The preceding discussion illustrates - the second significant distinction between chordal and external systems:

the chordal system is an actual, if approximate, volumetric flowmeter whereas the external system is a diametral velocimeter, Which places a greater burden on knowledge of the hydraulics at the location in which it is installed.

Incorporating a profile factor PF, in equation 17, the algorithm used by Caldon for chordal systems is obtained:

18) Q = (PF)(1D/2) {21 N (Wi Lw 2/ tan $,)(At. /(

t6)}

Where At,

= ta - tAe, tfABI tABI - 1non n.uid delay and tfAI taAJ - 'tnon fluid delay.

Transverse velocity components can affect chordal systems as they do external systems, but usually to a lesser degree. The vortices produced by a single bend 5 diameters upstream of a chordal UFM may affect the calibration by 0.1 or 0.2% (versus several percent for an external system without transverse velocity compensation). The swirl produced by nonplanar bends can significantly alter the calibration of both chordal and external systems unless the distance between the UFM and the second bend is enough to center the swirl. Generally speaking, UFMs are more forgiving of upstream and downstream hydraulics than turbine meters. By following a few rules, the use of flow conditioners can be avoided.

In chordal LEFMs, there is a pocket formed on the internal spool piece diameter by the aperture through which the acoustic beam passes as it makes its way

-- from the transducer well into the flow stream. If the transducer aperture is large with respect to the pipe internal diameter, the hydraulics and acoustics.of the pockets can influence the velocities measured. The profile factor for such installations, in addition to its other functions, must account for the influence of the pockets.

Summary of LEFM Principles The velocity measurements of Caldon's transit time ultrasonic systems rest on first principles. The accuracy with which one can measure velocity does not rest on an empirical relationship, but on the accuracy with which one can measure the transit

time, the dimensions, and, in the case of external systems, the acoustics of the installation.

Translating the velocity measured by an external UFM into a volumetric flow is essentially an empirical process. The calibration of external meters is sensitive to pipe condition and Reynolds Number, limiting their flow range in some applications.

The velocity measurements of Caldon's chordal UFMs lie along four mathematically specified, parallel chords. Because these four measurements are combined in accordance with the rules of a predictable numerical integration method the volumetric flow determination of a Caldon chordal system rests on first principles.

2 Handbook of Mathematical Functions, page 887, National Bureau of Standards, Applied Mathematics Series

A. Enclosure to letter, Susquehanna Benchmark Alarm Evaluation and Recommendations, Evaluation of Velocity Profile Change at SSES Unit 2, dated October 16, 2001 ER-262 Rev. 0 Count on Caldon Appendices

Evaluation of Velocity Profile Change at SSES Unit 2 Summary On October 6, 2001, a Profile Test (Benchmark Velocity) alarm occurred for the Loop A subsystem of the LEFM Check installed at Susquehanna Unit 2. This alarm occurs when the velocity measured on any one of the 4 paths, normalized to the average velocity and weighted according to its contribution to the total flow result, differs from a reference value by more than a preset amount (+

0.5% was the allowable deviation in weighted path velocity at the time of the alarm). A reference value for the velocity in each path was established at commissioning. The purpose of the alarm is to alert the user of the LEFM that the velocity profile may have changed from that which prevailed when the instrument's calibration was established.

When the alarm occurred, there was concern that the meter may have been malfunctioning. A review of the data shows, however, that the meter was performing exactly in accordance with its specifications and that, in fact, a significant profile change had occurred in Loop A. An evaluation of the profile data shows:

(1) The profile change was transient in nature, and (2) The (temporary) potential calibration error introduced by the profile change was no greater than about 0.1% and was in fact conservative. That is, the true flow was probably slightly lower than the indicated flow (by no more than 0.1% of reading) during the period when the profile was altered. [It should be noted that, because of the alarm, the plant was not using the LEFM to determine power, but, in accordance with its procedures, was using the venturi nozzles.]

In summary, this evaluation shows that the LEFM was operating within its design basis during the period when the Loop A profile differed from the reference. Because it appears possible that similar profile changes may occur again (see the discussion below), revised alarm settings will be implemented, to prevent these anticipated profile changes from causing the alarm in the future. The revised settings will still ensure that profile changes that could cause calibration errors larger than the design basis will be alerted.

Discussion The change in the velocity profile seen by the LEFM in the A Loop at SSES was probably produced by a decrease in the relative roughness of the upstream piping system. This decrease in roughness resulted in an increase in the swirl velocity seen by the Loop A LEFM. Swirl is typically produced by non planar changes in flow direction. The hydraulic geometries of loops A, B, and C in Susquehanna Unit 2 are very similar, but a swirl is present at the Loop A LEFM location, while none is present in Loop B or C. When the Loop A LEFM was commissioned, the tangential velocity of the swirl was modest-a tangential velocity of about +/- 4% of the axial velocity at the outside (short) paths (an 8%

difference in path velocities) and less than +/- 1% at the inside (long) paths. This pattern persisted for the months following commissioning.

HE011002 10/17/01 I

Attachment I

The change in profile that initiated the velocity alarm occurred on October 6, 200 1. On this date, a reduction in power to about 75% power appears to have brought about plant chemistry and/or flow changes that reduced the roughness in the feedwater piping upstream of the loop A LEFM. A reduction in roughness causes a flattening of the profile in and of itself, but for a plausible roughness change-say, a factor of 2-the amount of flattening would not be as great as the data show'.

However, a reduction in roughness also increases the velocity of the swirl at the LEFM location (because the rate of dissipation of the swirl in the straight pipe upstream of the LEFM is diminished).

The centripetal force produced by the high tangential velocity causes fluid traveling at high axial velocity to migrate to the outside of the pipe, further flattening the profile.

These changes can be seen in Figures 1A, 1B, and 1C. The change in axial velocity profile is characterized by the data plotted in Figure IA. The figure shows the ratio of the average short (outside) path velocity to the average long (inside) path velocity. A swirling (tangential) velocity component tends to add to the axial velocity component on paths on one side of the pipe centerline and subtract from the axial component on the other side. Hence the ratio of the average short path velocity to the average long path velocity measures what the axial profile would have been in the absence of swirl. It will be seen in.Figure IA that the axial profile flattens abruptly between 132 and 133 hour0.00154 days <br />0.0369 hours <br />2.199074e-4 weeks <br />5.06065e-5 months <br />s* --the ratio increases from roughly 0.87 to 0.89. This change is coincident with a reduction in power and feedwater flow to about 75% of rating (the velocity profile alarm occurred somewhat later, because of the long term averaging used in its implementation).

Simultaneously with the flattening of the profile, the swirl velocities on the short and long paths increase abruptly, as seen in Figures lB and 1C. These figures look at the normalized difference in the velocities measured by the outside paths and the inside paths. They indicate that the angular velocity of the swirl roughly doubled coincident with the down power. The swirl velocity is one half of the difference; Figure 1B indicates a swirl of about +/- 4% increasing to over +/- 7% in the outside paths The velocity profiles seen by the LEFMs in loops B and C show little or no change with the reduction in flow and power at 133 hours0.00154 days <br />0.0369 hours <br />2.199074e-4 weeks <br />5.06065e-5 months <br />. This can be seen from the data of Figures 2A and 3A. These profiles are more "round shouldered" than the profiles of loop A-their short-to-long path velocity ratios are about 0.83 versus 0.87 on loop A before the down power. This is probably because there is very little swirl present at these locations, as can be seen in Figures 2B and 3B. It is therefore not surprising that there is little change evident on these figures with the down power. [The velocity differences of the inside paths for loops B and C have not been plotted; they show smaller transverse velocity components than do the outside paths.]

Figures 1A, 1B and 1 C show the change in A loop profile brought about by the down power gradually disappearing in the hours following the return to full flow. This response suggests that the change in profile was caused by a change in wall roughness brought about by a water chemistry transient coincident with the down power. A change in feedwater chemistry is inherent with the

• A reduction in relative roughness from 0.0002 to 0.0001 would cause about half as much flattening as occurred on October 6.

"132 hours0.00153 days <br />0.0367 hours <br />2.18254e-4 weeks <br />5.0226e-5 months <br /> corresponds to 11:37 AM on October 6. The down power appears to begin an hour earlier.

HE011002 10/17/01 2

change in final feed temperature that accompanies a power reduction#. Additionally, heater drains, which can alter the dissolved and undissolved content of the feed, may be redirected during such transients". Changes in profile of the kind observed at Susquehanna have been seen in several other plants, and will be the subject of a Caldon Bulletin, to be issued in the near future.

It may be demonstrated that the (temporary) and limited flattening of the profile, as occurred during the transient of Figure 1, causes a 4 path LEFM to read conservatively by about 0.1%6*

The uncertainty analysis for the LEFM includes an allowance for profile factor (calibration) uncertainty that encompasses changes of this kind. Hence, the LEFM in Loop A at SSES was at all times operating within its design basis.

Changes to the velocity profile alarm settings for loop A should be implemented to prevent unnecessary alarms should such profile changes occur in the future. To select a revised profile test setpoint while retaining assurance that path velocity changes which could represent a profile outside the LEFM design basis would be alarmed, path velocities measured during calibration testing of the SSES spool pieces at Alden Research Labs were examined. These tests encompassed a several hydraulic geometries, including several orientations of the spools with respect to the upstream bend, and straight pipe. For each hydraulic geometry, the profile factor (calibration coefficient) for the spool-was measured, as well as the path velocities, over a range of flows. The data for the Loop A spool show that, over all hydraulic geometries, the span in the calibration coefficient was about 0.2%

(i.e., +/-0. 1%). Although the calibration remained nearly constant, the changes in geometry caused path velocity changes of as much as 3% on the inside (long) paths and 9 to 10% on the outside (short) paths. In computing the velocity change needed to initiate a profile alarm path velocities are weighted according to their contribution to the flow calculation. The weighting factors are, approximately, 0.11 for the short paths and 0.39 for the long paths. When the weighting factors are applied to the changes measured during calibration testing, a Profile Test alarm setting of at least 1.2% (more than twice the setting on October 6) is justified. This setting for the Profile Test alarm will provide the necessary protection without false actuations (the maximum weighted path velocity change seen in the transient of October 6 was only slightly above the setting at the time, 0.5%). To ensure that the profile protection is effective at or near plant rating, a setting for the profile alarm-enabling threshold of 90%

full flow is recommended. At lower flows, the LEFM will deliver a flow measurement accuracy of

+/- 0.4% of rating or better, even if weighted velocity changes greater than 1.2% occur. SSES calibration data, as well as other spool calibration data show that even extreme changes in profile are

1. Examination of the LEFM data through October 12 (beyond the range of the Figures) shows the gradual return continuing until a down power on October 12. When this occurred, the Loop A profile, which was still slightly flatter than originally, abruptly returned to its original shape. The response shows that down powers can lead to both smoothing and roughening of the loop A piping.

.. Plant personnel have suggested the following, plausible explanation: Reactor water level at SSES is controlled by changing the speed of the feed pumps in Loops A, B, and C. Different settings are employed for each of the feed pump governors-Loop A pump is the "lead" pump, while the pumps for Loops B and C are "followers". All small adjustments to flow are made by the A pump. This response was seen in the data of the October 6 transient; the change in flow in the A Loop was larger and more "busy" than either of the other loops. This control arrangement has prevailed since startup.

The constantly changing flow in A loop may be responsible for a corrosion layer having a different and smoother character than the other loops.

Calculation and experimental verification on file at Caldon. The theoretical maximum change for a fully developed profile at a Reynolds number of 3 x 107 is about 0.2%. That is, if the full developed profile suddenly became flat, the LEFM would read high by 0.2%.

HE011002 10/17/01 3

unlikely to cause calibration changes of more than 0.3 to 0.4% of reading. Hence, calorimetrics can be performed at all power levels below 90% with excellent accuracy, without the profile alarm.

Figure 1A Moter I Mwrt path long path rmto 0.976-0.06.

0.97

• -6 a

lIM 163 32M HE011002 10/17/01 4

Attachment I

Figure lB Metor I NORMAUZED OUTSIDE PATH OFFMENTIAL 0.10 0.14 0.12 nno Z

b.v

1S U-u (I (I 0.0 0.

0.0 0

0.0

~* 0.

N 0.0 0.

0.0 2

0n 1o0 160 bouws Figure 1C MteW I long path difserwn*

U4 25

.4b

• *.,*S.

02 4-.'

25 rll" 4 Sirltgt

. Sanest 0

a IM hour HE011002 10/17/01 5

Attachment I

Metr2 short path long path rallo iLO 0.83 6

6

'S~r1IS1 3,

Figures 2A and 2B M.otr 2 short pamt difforential n,.

t M

Im 50M 2!

" O M

a 4.LfUl I 01'

.SIrtgEl 4

A~ nP -

boun HEO11002 10/17/01 6

Attachment I

M3 short path long path ratio D.M.

0.956

'a 8.840 I

• rn. 0.84 D.B3 n.8R' 0.02.

0.0 16 0.014 IL S001 q.DD U.GM 0.00D a.002 0

howm Figures 3A and 3B W short Path. diumnial m,ba p.*.d*..

4

& Se~I2e1 a

howr BEO11002 10/17/01 7

Attachment I

B. Letter, H. Estrada, Caldon to Ms. Debra Echols, Tennessee Valley Authority, dated September 7, 2001,"Change in Velocity Profile Measured by the WBN LEFM

[Check]"

ER-262 Rev. 0 Count on Caldon Appendices

mo u

0 Caldon, Inc.

September 7, 2001 Ms. Debra Echols (for distribution)

Tennessee Valley Authority Watts Bar Nuclear Power Station Subject Change in Velocity Profile Measured by the WBN LEFM Check

Dear Ms. Echols:

This letter provides Caldon's evaluation of the effect, on the accuracy of the LEFM Check, due to the change in the fluid velocity profile recently seen by this instrument The change in profile was observed following restart after a plant trip, and was sufficient to trigger the LEFM Check velocity profile alarm. The alarm is intended to alert users of the LEFM Check that the velocity profile has changed significantly from that measured at the Instruments commissioning. The profile measured at commissioning is, In turn, compared with that measured during calibration testing of the LEFM Check, to ensure applicability of the calibration in the field. It is Caldon's practice, when a user reports a profile alarm, to evaluate the specifics of the change, to ensure that the calibration for the meter still applies and that its uncertainty is within its design basis. It should be noted that profile alarms are unusual, but have occurred in 2 or 3 chordal systems currently in service.

The LEFM Check at Watts Bar is installed in a 32 inch header about 45 diameters downstream of a single 90 0 bend.

High pressure feedwater heaters feed the header upstream of the bend. The velocity profile data for Watts Bar, recorded before the plant trip and following the profile alarm are given in the table below. Velocities are normalized to the velocity averaged over the pipe cross section. V1 and V4 are the velocities measured along the two outside' (short) chords of the LEFM Check; V2 and V3 are measured along the two inside (long) chords.

Vl V2 V3 V4 VSHORT/VLONG (average)

Profile before plant trip 0.86 1.03 1.04 0.90 0.85 Profile with alarm 0.82 1.00 1.05 1.01 0.89 The profile before the trip is typical of developed flow in a straight pipe. The slight asymmetry in the profile before the trip (V3 and V4 are slightly larger than V2 and V1) is believed to be due to a very small swid residual from the interaction of the velocity profile distortion produced by the heater discharge lines and the bend upstream of the LEFM Check.

1070 Banksville Avenue

• Pittsburgh, PA 15216 Tel: 412-341-9920 - Fax: 412-341-9951 - Web: www.caldon.net

The swirl has increased following the trip, based on the increased asymmetry of V3 and V4 versus V1 and V2, though it is still small (about 9% of the axial velocity near the outer pipe wall). The swirl is centered in both cases and produces no error in the LEFM Check reading.

The overall shape of the profile following the trip is flatter than it was before the trip. This isthe reason that the ratio of the average short path velocities to the average long path velocities increases from 0.85 to 0.89. A profile of this short path/long path ratio is not unusual, but is characteristic of developed flow at high Reynolds Number in very smooth pipe. It appears that the trip, and the subsequent operation of the feedwater system removed some or most of the rough corrosion film from the 45 diameters of pipe upstream of the LEFM Check, thereby producing a flatter profile and reducing the rate at which the swirl produced by the bend is dissipated. It is understood that condenser vacuum was maintained during the shutdown and the feedwater system was operated in a "long recycle" configuration throughout the period. This operating history, coupled by the sudden temperature change inherent in the shutdown, is consistent with the scale removal hypothesis.

The flatter profile does not significantly change the calibration of the LEFM Check, nor does it change the uncertainties associated with the calibration. In fact, the present meter factor is likely to be slightly conservative (less than 0.1%). Accordingly, we recommend that operation using the LEFM Check for thermal power computations be resumed. Because the change in profile is likely to persist for a long period-the rough film will likely take months or years to reform, if it reforms at all-we recommend that the settings of the velocity profile alarm be revised. Data for these revised settings will be provided under separate cover.

Sincerely Herb Estrada Chief Engineer Cc: Emie Hauser Cal Hastings Don Augenstein Ed Madera Ryan Hannas 1070 Banksville Avenue - Pittsburgh, PA 15216 Tel: 412-341-9920 - Fax: 412-341-9951

• Web: www.caldon.net

C. Calculation: Determination of Axial Velocity Profiles from Chordal Velocity Measurements, dated October 31, 2001.

ER-262 Rev. 0 Count on Caldon Appendices

ER-262 APPENDIX C JANUARY 2002 REV 0 CALDON, INC.

ENGINEERING REPORT: ER-262 APPENDIX C CALCULATION DETERMINATION OF AXIAL VELOCITY PROFILES FROM CHORDAL VELOCITY MEASUREMENTS Prepared By: Herb Estrada Reviewed By: Emie Hauser ER-262R0 Appendix C Count on Caldon I

0 r..v 0

'3 Calculation Determination of Axial Velocity Profiles from Chordal Velocity Measurements A. Purpose The purpose of this calculation is to describe the methodology whereby the velocity measurements of 4 path chordal transit time flowmeters in a specific hydraulic geometry can be used to determine the mean velocity along a diametral path in that same hydraulic geometry. The calculation also describes how these data can be used to compute calibration coefficients for 4 path chordal systems and for external (diametral path) systems.

B. Assumptions

1. Any swirl that may be present is centered. The 4 paths of a chordal system (two long, inside paths and two short, outside paths) are parallel to each other and are symmetrical with respect to the pipe centerline. When the swirl is centered, the swirl (tangential) velocity projections on each of the two acoustic paths on one side of the centerline are equal and opposite to the components projected onto the two acoustic paths on the other side of the centerline. The contribution to the path velocity readings can be determined from the difference in path velocities, and the axial profile shape can be determined by averaging the velocities measured on inner chords and the velocities measured on outer chords. Experimental data indicate that the centripetal forces associated with swirling flow tend to center the swirl in about 15 diameters of straight pipe.1 Furthermore, Caldon practice is to orient the acoustic paths normal to the plane of the last bend, which orientation leads to a symmetrical profile in even shorter lengths (about 5 diameters).2

2. Axial velocity profiles at chordal flowmeter locations can be characterized by the ratio of the measured axial short path (outside chord) velocity to the average long2path (inside chord) velocity (i.e., the swirl contribution has been removed). From these data the velocity as a function of local radius over the pipe cross section can be fitted using the inverse power law by varying the exponent.

The justification for this procedure is based on the work of Nikuradse and others on flow in smooth

.3 and rough pipe.

C. Summary Figure 1 presents the relationship between the profile factor for a 4 chord (4 path) ultrasonic transit time system, calculated using an inverse power law fit of short and long path velocities, and the ratio of average short path velocity to average long path velocity (SP/LP VR).

Figure 2 presents the relationship between the profile factor for a single (diametral) path ultrasonic system, also calculated using an inverse power law fit of short and long path velocities, and the ratio of average short path velocity to average long path velocity (SP/LP VR).

1 Murakami et al, Studies on Fluid Flow in Three Dimensional Bend Conduits, JSME Bulletin, Vol. 12, No. 54, December 1969 2 Westinghouse Oceanic Division Report OEM 78-40, February 1979, G.P. Erickson and P.G. Spink 3 Boundary Layer Theory, Dr. H. Schlichting, McGraw Hill, Sixth Edition, Chapters XIX and XX ER-262R0 Appendix C Count on Caldon 2

Table 1 provides average short path velocity to average long path velocity ratios (SP/LP VRs) characterizing the variations in chordal path data measured at 18 chordal installations. The Table also includes the calculated variations in calibration (Profile Factor) for 4 chord systems and diametral path systems experiencing the profile variations tabulated. The calculated calibration variations are based on linear fits of the curves of Figures 1 and 2.

D. Calculation

1. Symmetrical axial profiles can be described using the so called inverse power law which represents the spatial axial velocity distribution in apipe of circular cross section as follows:

u / U = (y / R) Vn Where u is local fluid velocity, U is the fluid velocity at the centerline, y is the distance from the pipe wall, R is the internal radius of the pipe, and n is an empirically determined exponent.

The inverse power law was used extensively by Nikuradse and others to fit flow profiles over a wide range of Reynolds Numbers in rough and smooth pipe, in the development of the methodology for calculating friction losses in turbulent flow4.

2. The mean axial velocity through the pipe (i.e., the local axial velocity averaged over the pipe cross section) is given by:

uAvG = f U (r) dA/JfA Here the local radius, r = R - y, and The incremental area, dA = 2ir dr Using the relationship of paragraph 1 and writing the integral in terms of y UAVO = - (U!/ R2) f (y/ R) "n x 2n (R -y) dy Where the integration is performed from R to zero.

This integration yields the following relationship between the mean axial velocity uAvc and the centerline velocity U:

U = UAV

[1I + 1.5/n+ 0.5/n 2]

For a given n, then, the centerline velocity can be computed from the expression above.

4 Boundary Layer Theory, op. cit.

ER-262R0 Appendix C Count on Caldon 3

A selection of n also allows the computation of the mean velocity along any chordal path within the pipe. Rectilinear coordinates will be employed. The x axis will be defined as parallel to the chord and passing through the pipe centerline. The y axis will be defined as perpendicular to the chord and passing through the pipe centerline. (NOTE: The coordinate y does not correspond to the variable of integration in paragraphs 1 and 2.] The y coordinate defines the specific chordal location relative to a centerplane defined by the x axis and the axial centerline of the pipe. Three specific y coordinates are of interest:

For the short (outside) chords in Gaussian quadrature integration using Legendre spacing, yj =

0.861R For the long (inside) chords in Gaussian quadrature integration using Legendre spacing, y2 = 0.340R

• For the diametral chord inherent in any externally mounted ultrasonic meter, y3 =.OOR At any location, x, along the chord at yi a local radius, r can be computed:

r = [x 2 + yi 2 ]1/2 For the selected n, the local velocity u ( r ) at this location can then be computed using the relation of paragraph 1 u (r)= U (1-r / R) 1 The mean velocity measured at any chord is:

UCHORD = S u (x, yi) dx / S dx This integration is performed numerically by dividing the chord length into increments Ax.

Increments of 0.001 of the chord length X were used. Here X = [R 2 - y?2 ]V/2 Note that the integration process is carried out over only half of the total chordal length. That is, it is performed from 0 to X; the chord extends from -X to + X. However, because the profile is symmetrical about 0, the integration as performed gets the correct result.

3. The calculation described in the preceding paragraph has been performed using an Excel spreadsheet 5. The process is as follows:

• An exponent n is assumed. (Profiles for values of n ranging from 6 to 30 were calculated).

" The centerline velocity is computed relative to a mean velocity of 1.00.

• For chords located at each of the three y coordinates of interest, the mean axial velocity for the chord is calculated. In each case the procedure is:

5 The spreadsheet is on file at Caldon.

ER-262R0 Appendix C Count on Caldon

Starting at x = 0, u (x, yi) is calculated.

x is incremented by an amount Ax = Xi / 1000)

The value of u (x, y,) Ax is computed The cumulative sum of u (x, yi) Ax is computed.

The process is continued until x =X.

The mean velocity along the chord is obtained by dividing the cumulative sum of u (x, y,) Ax byX The ratio of the mean long path to mean short path velocity that would be measured by a 4 path chordal system, with a profile as defined by the assumed exponent n, is calculated.

The theoretical profile factors (calibration coefficients) for a 4 path chordal system and a diametral (external) system, operating in the velocity profile characterized by the exponent n, are computed.

The procedures for these calculations are described below.

4. A Profile Factor (PF) as used in Caldon instruments is defined as the quotient of the true flow to the flow as measured by the instrument prior to any correction. Hence, PF = (uTuE ARuE ) / (Um..s A~mAs)

Here uTRuE is the true mean axial velocity over-the pipe cross section, AruE is the exact area of the pipe cross section, umEAs is the axial velocity measured by the instrument, and AmEps is the cross sectional area embedded in the measurement of the instrument.

This analysis will assume no errors in the area measurements.

5. Accordingly, the Profile Factor, PF1 for a diametral path (external) system is given by PFI = (urME) / (UMES) = 1 / u*.*s u U (x, 0.0) dx / R]

Where the integration is performed from 0 to R

6. For a 4 path chordal system, the measured mean short chord velocity, USHORT, is multiplied by a factor kSHORT that reflects the weighting specified for this chord by the quadrature integration "

method and the chord length. Likewise the mean long chord velocity ULONG is weighted by a factor kLONG that reflects the weighting specified for this chord by the quadrature integration method and the chord length. Thus, the Profile Factor for a 4 path chordal system, PF4, is given by PF4 = I/[ 2 x kSHORT USHORT + 2 x kLONG ULONG]

Where kSHORT = 0.11 2, kLONG = 0.388, USHORT = f u (x, 0.86R) dx / XSHORT, and ULONG = f U (X, 0.34R) dx / XLONG ER-262R0 Appendix C Count on Caldon 5

7. As previously noted, mean velocities for the short chords, the long chords, and the diameter were calculated for profiles whose inverse exponent n ranged from 6 to 30. Profile factors for the 4 chord and diametral systems were also calculated. For each selected exponent, the profile factors for both systems were then plotted against the ratio of the short path velocity to the long path velocity (SP/LP VR) for that exponent. The Profile Factor (calibration coefficient) for a 4 chord system is graphed against SP/LP VR in Figure 1. A linear fit (shown in the figure) has been used to characterize the relationship. The Profile Factor (calibration coefficient) for a diametral (external) system is graphed against SP/LP VR in Figure 2. Again, a linear fit (shown in the figure) has been used to characterize this relationship. For comparative purposes Figure 2 also shows the 4 chord system Profile Factor (the flatter curve near the top).

The linear fits of the Profile Factor relations are as follows:

• PF1 = 0.368 (SP/LP VR) + 0.6331

• PF4 =- 0.0167 (SP/LP VR)+ 1.0167 These relations have been used to calculate the calibration changes that variations in the short and long path velocities measured in 18 Caldon chordal systems would produce in diametral and 4 chord systems. Results are tabulated in Table 1.

ER-262RO Appendix C Count on Caldon 6

Figure 1 Profile Factor 4 path chordal system vs. SP/LP VR 1.00500 1.00450-1.00400 1.00350 1.00300 u 1.00250 1.00200 1.00150 Y -0.01 66x + 1.0167 1.00100 1.00050 1.00000 1 0.75000 0.80000 0.85000 0.90000 0.95000 1.00000 SP/LP VR ER-262R0 Appendix C Count on Caldon 7

Lao~

0---

a I Q Figure 2 Calibration Coefficient (PF) versus short chordllong chord velocity ratio 1.01 1.00-0.99-0.98-0.97-

"0.96 0.95-0.94-0.93.-

0.92 0.91 0.75 0.80 0.85 0.90 0.95 short path/long path velocity ratio diametral path --*-4path Gauss -

Linear (4path Gauss) -

Linear (diametral path) 1.00 ER-262R0 Appendix C Count on Caldon 8

~o 0 ~Q Table 1 Calculated 4 Path and Single Path Profile Factors* versus Measured Chordal Velocity Ratios Based on a random sample of logged data over periods of operation ranging from 2 months to several years Plant/Unit Hydraulic Geometry Max Min 4Path Chordal PF Diametral Path PF SP/LP SP/LP vie VW Max Min A

Max Min A

WBN 1 LEFM Check 45D downstream 0.892 0.854 1.0024 1.0018 0.0006 0.961 0.947 0.014 of single 90 0 bend. 3 HP heater feeds upstream of bend include non planar reverse bend SSES 2 Loop A Three loops similar. LEFM 0.894 0.864 1.0023 1.0018 0.0005 0.962 0.951 0.011 Loop B Check -13D downstream of 0.837 0.827 1.0029 1.0027 0.0002 0.941 0.937 0.004 Loop C single 90 0 bend. Non planar 90 (

0.830 0.822 1.0030 1.0028 0.0001 0.939 0.936 0.003 bend 11 to 12 diameters upstream.

IP 2 Loop 21 LEFM in each loop between 10 0.894 0.884 1.0019 1.0018 0.0002 0.962 0.958 0.004 Loop 22 and 15D downstream of 90 0 0.931 0.883 1.0020 1.0012 0.0008 0.976 0.958 0.018 Loop 23 bend with nonplanar 90 0 bend 0.916 0.874.

1.0021 1.0014 0.0007 0.970 0.955 0.015 Loop 24 10D upstream 0.939 0.917 1.0014 1.0010 0.0004 0.979 0.971 0.008 IP 3 Loop 31 LEFM in each loop 6D 0.940 0.921 1.0013 1.0010 0.0003 0.979 0.972 0.007 Loop 32 downstream of 90 0 bend with 0.925 0.916 1.0014 1.0012 0.0002 0.974 0.970 0.004 Loop 33 nonplanar 90 0 bend 1OD 0.952 0.932 1.0011 1.0008 0.0003 0.983 0.976 0.007 Loop 34 upstream 0.976 0.952 1:0008 1.0004 0.0004 0.992 0.983 0.009 CP 1

-LEFM in each unit I ID 0.918 0.914 1.0014 1.0014 0.0000 0.971 0.969 0.002 CP2 downstream of 90 0 bend Non 0.909 0.908 1.0015 1.0015 0.0000 0.967 0.967 0.000 planar feed - 18 diameters upstream.

Continued, next page ER-262R0 Appendix C Count on Caldon 9

~O0 a El Table 1, continued PlantlUnit Hydraulic Geometry Max Min 4Path Chordal PF Diametral Path PF SP/LP SP/LP Max Min A

Max Min A

VR VR PI 2 Loop 31 LEFM in each loop--20D 0.867 0.851 1.0023 1.0020 0.0003 0.957 0.951 0.006 Loop 32 downstream of 90 0 bend. Each 0.881 0.868 1.0022 1.0020 0.0002 0.957 0.953 0.004 loop is fed from the branches of a non planar symmetrical lateral 4 diameters upstream of bends..

BV 1 UI LEFM --6 D downstream of 0.922 0.913 1.0015 1.0013 0.0002 0.972 0.969 0.003 BV 2 header, 2 non planar feeds 0.920 0.915 1.0014 1.0013 0.0001 0.972 0.970 0.002 upstream (UI)

U2 LEFM -10 D downstream of header, 2 non planar feeds upstream (U1)

Mean High - Low PF (A),

0.0003 0.007 1 o (standard deviation)

+/-0.0002

+/-0.005 Average Diametral Path PF: 0.964 Notes

• A Profile Factor is the calibration coefficient for an ultrasonic meter. It is sometimes referred to as a "velocity profile correction factor" and is equivalent to the discharge coefficient of a flow nozzle.

+ SP/LP VR is the ratio of the average velocity projected onto the short chords (or paths) to the average velocity projected onto the long chords.

• • A Profile Factor of 0.953, based on model tests, was employed on an external (Diametral Path) ultrasonic meter installed 20D upstream of the LEFM Check (i.e.,

25D downstream of the bend).

• • • The indication of the external meter installed at 25 diameters downstream of the bend shifted about 1.6% relative to the indication of the 4 path chordal instrument during an operational sequence when the chordal velocity ratio changed from its minimum to its maximum value. Allowing for a change in the calibration of the 4 path meter of 0.06%, the net calibration change measured for the external meter at 25D was about 1.5%, a figure entirely consistent with the 1.4% calculated from the change in the measured chordal velocities.

ER-262R0 Appendix C Count on Caldon 10

D. Summary Table: Evaluation of Hydraulic Configurations and Uncertainties for Operating External LEFMs ER-262 Rev. 0 Count on Caldon Appendices

ER-262 APPENDIX D:

SUMMARY

TABLE:

EVALUATION OF HYDRAULIC CONFIGURATIONS AND UNCERTAINTIES FOR OPERATING EXTERNAL LEFMS

('."

Summary Table: Evaluation of Hydraulic Configurations and Uncertainties for Operating External LEFMs Results of Caldon's analysis indicate that current external meter applications in the industry fall into one of four categories:

A. No measurable effect. The LEFM 8300 external meter is installed downstream of and in close proximity to a flow straightener designed to dominate the local velocity profile. This effectively isolates the LEFM from effects of changing upstream velocity profiles.

B. Possible effect modeled and bounded. Potential velocity profile changes at the installation location were modeled and are bounded by calibration testing.

C. Possible effect bounded. The calibration testing did not specifically address the profile changes that have since been observed. However, their effect on meter accuracy is bounded by the existing uncertainty allowance.

D. Uncertainty bounds affected. The calibration testing did not specifically address the profile changes since observed. Furthermore, their effect on meter accuracy is not bounded by the existing uncertainty allowance.

No action is necessary for any of these categories except category D.

All LEFM 8300 installations were evaluated. As shown by the following table, only one of the 55 feedwater pipes with LEFM 8300 external meters falls in category D.

Plant Category Report Cofrentes A

ER-236 Fitz Patrick A

ER-238 Kashiwazaki Unit 1 A

ER-239 Kashiwazaki Unit 5 A

ER-241 Perry A

ER-242 River Bend A

ER-244 Doel Units 3 and 4 B

ER-228 Grand Gulf B

ER-229 Millstone Unit 3 B

ER-230 Nine Mile Point I B

ER-231 Nine Mile Point 2 B

ER-232 Palo Verde Units 1, 2, and 3 B

ER-233 Trillo Unit 1 B

ER-234 Vandellos Unit 2 B

ER-235 Doel Units 1 and 2 B

ER-237 Kashiwazaki Unit 4 B

ER-240 VC Summer B

ER-247 St. Lucie Unit 2 Loop A = B Loop B = C ER-246 Quad Cities Units I and 2 C

ER-243 Sequoyah Units I and 2 C

ER-245 Watts Bar D

ER-250

E. Scoping Calculation: Errors in Flow Nozzles with Swirl Velocity of 10% Axial Velocity ER-262 Rev. 0 Count on Caldon Appendices

ER-262 APPENDIX E:

SCOPING CALCULATION:

ERRORS IN FLOW NOZZLES WITH SWIRL VELOCITY OF 10% AXIAL VELOCITY ER-262 Appendix E Count on Caldon I

Scoping Calculation:

Errors in Flow Nozzles with Swirl Velocity of 10% Axial Velocity

Purpose:

T'he purpose of this calculation is to provide an approximate estimate of the error in the flow measurement of a nozzle, produced by swirl having a tangential velocity of 10% of the axial velocity.

Errors will be calculated for nozzles having beta (diameter) ratios of 0.5 and 0.7.

Assumptions:

I. The hydraulic losses between the upstream (pipe) tap of the nozzle based flow measurement and the throat tap are negligible. That is, the total pressure at these two stations is the same.

.2. The flow is incompressible. That is, the product of 'the mean axial velocity and the cross sectional area at the upstream tap location equals the product of the mean axial velocity and the cross sectional area at the throat tap location.

3. The swirl can be characterized as a rotating disk of fluid, having a tangential velocity at the pipe wall equal to the product of the radius and the angular velocity.

4. Rotational momentum is conserved between the upstream pipe tap and the throat tap. That is, the products of the rotational moment of inertia and the angular velocity of the fluid at each of these stations are equal.

Summary:

With a tangential velocity due to swirl of 10% of the axial velocity, a flow nozzle with a beta ratio of 0.5 will read in error by 2%. The actual flow will be less than the indicated flow.

This same tangential velocity will produce an error of 0.65% in a flow nozzle having a beta ratio of 0.7.

Again the actual flow will be less than the indicated flow.

ER-262 Appendix ECononado2 Count on Caldon 2

Calculation:

1. The nozzle configuration and nomenclature are shown in the sketch below Station 1

2 Total pressure PrT Pr2 Static pressure psi Ps2 Axial Velocity VI V2 Area Ai A2 Internal Radius R,

R2 Moment of Inertia I 12 Angular Velocity o &2

2. The fluid energy per unit volume at each station is given by the total pressure. In accordance with Assumption 1:

prI = (potential energy/ unit volume + kinetic energy/ unit volume),

PT2 = (potential energy/ unit volume + kinetic energy/ unit volume) 2

3. The static pressure defines the potential energy/ unit volume at each station. Rearranging terms in the above equations and noting the difference in total pressure is zero, the difference in static pressures is given by Psi - ps2 = (kinetic energy/ unit volume) 2 - (kinetic energy/ unit volume)I

4. In the base case no swirl is present. In this case, the difference in kinetic energy per unit volume is given by:

PSI - ps2 = / P V2 g -

p V12/g where g is the gravitational constant.

5. The velocity at station 2 is determined in terms of the velocity at station 1 using Assumption 2.

V, A, =V 2 A2 V2 = Vi Ai/A 2 = Vi Ri2/R 22 = V, / p2 ER-262 Appendix E Count on Caldon 3

The term P3 is defined as the ratio of the throat diameter to the pipe diameter. Hence P3 equals the ratio of the throat radius to the pipe radius.

6. Substituting for V2 in the equation of paragraph 4, the differential pressure for the nozzle is given by psI - ps 2 = Ap = '(p / g) (VI /

2 )2 (p/g)V 2

(/g)

V2 [ (1/

4)_1]

7. For the case where swirl is present, rotational kinetic energy per unit volume must be added to the kinetic energy per unit volume term. Using Assumption 3, the rotational kinetic energy per unit volume, KERN at any station is given by KERN = M (I

)/ AAL Where AL is a unit of axial length The rotational moment of inertia of a rotating disc of thickness AL is given by' I = (p / g) (t R4/ 4) AL The term AAL is given by AAL =t Rý AL Hence KERN = M/ (p / g) (le2 2 / 4)

8. Assumption 4 implies that (0o))

(Io)2 Using the equation for moment of inertia from paragraph 7 in this equation, and canceling common terms R14 oi= R24 Thus co

(R1 / R2) 4 = o* (1/ 34)

9. At each station, the rotational kinetic energy per unit volume adds to the kinetic energy due to the axial velocity. It therefore increases the difference in static pressures by an amount equal to the difference between the rotational kinetic energy per unit volume terms at stations 1 and 2. The net error in the pressure differential 8Ap is 8Ap = (KER/V) 2 -(KER/V), = 2 (p / g) (R,2 Q2 / 4)[ (1/ P3) -1]

Eshbach, Handbook of Engineering Fundamentals First Edition Chapter 4 ER-262 Appendix E Count on Caldon 4

In the absence of swirl, the differential pressure for the nozzle was derived in paragraph 6:

Ap = V2 (p / g) VI2[(1' 4)-I Hence the per unit error in differential pressure, EAp is the quotient of these expressions.

EAP = {(R12 Q*2 14[(l 3)-]I{V12 [ (I/ p*4)_I]

Noting that R, co, is the tangential velocity at station 1, VTI, the per unit pressure error is E~p = 1/4/ (VT I / VI)2 [ (l/1 p) _11}/ [ (i/ p4) _ 1]

10. Since volumetric flow is proportional to velocity, and differential pressure is proportional to the square of velocity, the per unit error in flow, 8Q/Q is one-half the per unit error in pressure.

Accordingly, for a tangential velocity of 10% of the mean axial velocity 8Q/Q = ' EAp = 1/8 (0.1)2 [ (I/

l) -l]}/[ (p/f34) - 1]

For P = 0.5, 8Q/Q = 2.0%

For 3 = 0.7, 8Q/Q = 0.65%

Note that in both cases the swirl causes the nozzle's flow indication to be high, since the rotational kinetic energy increases the differential pressure for a given axial velocity.

ER-262 Appendix E Count on Caldon 5

F. Plant Data, 4 and 8 Path Chordal Installations ER-262 Rev. 0 Count on Caldon Appendices

TABLE OF CONTENTS

1.

Plant Data Watts Bar Unit 1

2.

Plant Data Susquehanna Unit 2 Loops A, B, and C

3.

Plant Data Indian Point Unit 2 Loops 21, 22, 23, and 24

4.

Plant Data Indian Point Unit 3 Loops 31, 32, 33, and 34

5.

Plant Data Comanche Peak Unit-1 and Comanche Peak Unit 2

6.

Plant Data Prairie Island Unit 2 Loop A and B

7.

Plant Data Beaver Valley Unit 1 and Beaver Valley Unit 2 ER-262 Rev. 0 ER-262 Rev.

0 Count on Caldon Apni Appendix F

Plant Name:

Watts Bar Unit I Feedwater Measurement System:

LEFM,/

Installation Geometry:

45 L/D Downstream of Single 900 Elbow Chordal Meter Measurement Error= 0.06%

0 hrtrc of Radius ER-262 Rev. 0 Count on Caldon Appendix F

cZ~7E~ 0 L~c~E~ o ER-262 Rev. 0 Count on Caldon Appendix F

Unit 1 02:46:21 2001/08/29 Configuration Files AIARM.INI FAT.INI HYDRAULI.INI METER.INI PARAMETR.INI P CONFIG.INI PROPERTY.INI SETUP.INI 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 18:15:40 18:15:40 18:15:40 18: 15: 40 18: 15: 40 18: 15: 40 18:15:40 18: 15: 40 18:15:40 18:15:40 18:15:40 18: 15: 40 18:15:40 18:15:40 18: 15:40 18:25:40 FFFED282 FFFFEB2F FFFF4541 FFFD66BF FFFBSAE0 FFFF82DC FFFF6C54 FFFF9D29 FFF89717 FFF899D5 FFF899D5 FFF899D5 FFF899D5 FFF899D5 FFF899D5 FFF899D5 I

)

I Setup Files Setapul.txt Setapu2.txt Setapu3.txt Setapu4.txt Setapu5.txt Setapu6.txt Setapu7.txt Setapu8.txt I

I Unit Unit Unit Unit 11 1

1 Current Average Maximum Minimum Flow:

Flow:

Flow:

Flow:

Unit 1 Deviation Flow:

I 3

Unit Unit Unit Unit Unit 11 1

1 Current Temp:

Average Temp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

82.50 82.39 82.88 81.91 0.18 443.7 443.7 443.9 443.6 0:0 ALERT ALERT Unit 1 Current System Status:

Unit 1 Minimum System Status:

I Unit Unit Unit Unit Unit 11 1

1 1

Current Mass Flow:

Average Mass Flow:

Maximum Mass Flow:

Minimum Mass Flow:

Deviation Mass Flow:

15463.292 15442.563 15532.904 15350.739 34.223 Unit 1 Uncertainty:

I Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1

1 1

1 1

1 1

1 1

1 Current Flow:

Average Flow:

Maximum Flow:

Minimum Flow:

Deviation Flow:

Current Temp:

Average Temp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

0.11 82.50 82.39 82.88 81.91 0.18 443.7 443.7 443.9 443.6 0.0 J

Meter 1 Current Press:

1159.77 Meter 1 Average Press:

1158.10 Meter I Maximum Press:

1160.50 Meter 1 Minimum Press:

1155.75 Meter 1 Deviation Press:

0.04 Meter 1 Current Meter Status:

ALERT Meter 1 Minimum Meter Status:

ALERT Meter I Current Mass Flow:

15463.292 Meter 1 Average Mass Flow:

15442.563 Meter 1 Maximum Mass Flow:

15532.904 Meter 1 Minimum Mass Flow:

15350.739 Meter 1 Deviation Mass Flow:

34.223 Meter 1 Uncertainty:

0.11 Path 1 Path 2 Path 3 Path 4 Meter I Current Variance:

10167.92 19972.27 14771.31 8568.18 Meter 1 Average Vnorm:

0.8648 1.0277 1.0402 0.8996 Meter 1 Current Vnorm:

0.8679 1.0281 1.0408 A0.8933 Meter 1 Maximum Vnorm:

0.8831 1.0395 1.0528 0.9166 Meter 1 Minimum Vnorm:

0.8484 1.0151 1.0288 0.8772 Meter 1 Deviation Vnorm:

0.006 0.004 0.004 0.006 Meter 1 Benchmark Vnorm:

0.8648 1.0277 1.0402 0.8995 Meter 1 Limit % Vnorm:

0.50 0.50 0.50 0.50 Meter 1 Average Gain:

66.01 70.39 76.07 66.04 Meter 1 Current Gain:

66.01 70.41 76.13 65.97 Meter 1 Maximum Gain:

66.33 70.68 76.37 66.25 Meter I Minimum Gain:

65.66 70.17 75.78 65.82 Meter 1 Deviation Gain:

0.09 0.08 0.09 0.07 Meter 1 Limit Gain:

76.00 76.00 76.00 76.00 Meter 1 Current Gain Up:

65.54 70.09 76.21 65.39 Meter 1 Current Gain Down:

66.33 70.56 75.90 66.48 Meter 1 Current TPGain Up:

70.72 70.72 70.56 70.56 Meter 1 Current TPGain Down:

70.56 70.56 70.56 70.56 Meter 1 Average S/N Ratio:

38.50 26.71 15.31

.38.33 Meter 1 Current S/N Ratio:

39.73 27.35 15.27 39.49 Meter 1 MaximumS/N Ratio:

40.66 29.16 17.02 40.97 Meter 1 Minimum S/N Ratio:

35.50 23.83 13.54 35.28 Meter 1 Deviation S/N Ratio:

1.47 1.28 0.70 1.60 Meter 1 Average TDown:

478419 823170 823193 478446 Meter 1 Current TDown:

478373 823095 823121 478413 Meter 1 Maximum TDown:

478533 823378 823398 478567 Meter 1 Minimum TDown:

478325 823008 823010 478339 Meter 1 Deviation TDown:

35 60 61 36 Meter 1 Current TPTDown:

4000747 4000748 4000746 4000747 Meter 1 Average DeltaT:

2158.4 4812.0 4861.4 2209.0 Meter 1 Current DeltaT:

2168.5 4819.0 4869.7 2196.0 Meter 1 Maximum DeltaT:

2207.3 4876.3 4919.1 2254.9 Meter 1 Minimum DeltaT:

2107.8 4746.0 4794.5 2160.4 Meter 1 Deviation DeltaT:

15.7 22.8 23.5 15.4

Meter 1 Current TPDeltaT:

Meter 1 Current Path Status:

Meter 1 Minimum Path Status:

-2.3

-0.6

-0.6

-1.4 i

Meter Meter Meter Meter Meter Meter 1

1 1

1 1

1 Average Reject %

Current Reject %

Maximum Reject %

Minimum Reject %:

Deviation Reject %

Incoming Samples:

NORMAL NORMAL 0.1 0.0 2.8 0.0 0.3

'719 0

NORMAL NORMAL 0.1 0.0 1.2 0.0 0.2 719 0

ALERT ALERT 2.0 3.5 6.5 0.0 1.0 719 0

NORMAL NORMAL 0.0 0.0 1.5 0.0 0.2 719 0

I Meter I Number Failed Rejects:

I I

J I

Alarm Log Events 2001/08/29 01:46:18 2001/08/29 01:46:18 2001/08/29 01:46:19 2001/08/29 01:46:19 2001/08/29 01:46:33 2001/08/29 01:46:33 2001/08/29 01:46:34 2001/08/29 01:46:34 2001/08/29 01:47:08 2001/08/29 01:47:08 2001/08/29 01:47:09 2001/08/29 01:47:09 2001/08/29 01:47:28 2001/08/29 01:47:28 2001/08/29 01:47:29 2001/08/29 01:47:29 2001/08/29 01:47:48 2001/08/29 01:47:48 2001/08/29 01:47:49 2001/08/29 01:47:49 2001/08/29 01:47:53 2001/08/29.01:47:53 2001/08/29 01:47:54 2001/08/29 01:47:54 2001/08/29 01:48:03 2001/08/29 01:48:03 2001/08/29 01:48:04 2001/08/29 01:48:04 2001/08/29 01:48:08 2001/08/29 01:48:08 2001/08/29 01:48:09 2001/08/29 01:48:09 2001/08/29 01:48:13 2001/08/29 01:48:13 2001/08/29 01:48:14 2001/08/29 01:48:14 2001/08/29 01:48:23 2001/08/29 01:48:23 2001/08/29 01:48:24 2001/08/29 01:48:24 2001/08/29 01:48:28 2001/08/29 01:48:28 Meter 1 ALERT Unit 1 ALERT Meter 1 Path.3 Alert --

Gain Meter 1 Path 3 ALERT Meter I NOP14AL Unit 1 NORMAL Meter 1 Path 3 Pass --

Gain Meter 1 Path 3 NORMAL Meter 1 ALERT Unit 1 ALERT Meter 1 Path 3 Alert --

Gain Meter 1 Path 3 ALERT Meter 1 NORMAL Unit 1 NORMAL Meter 1 Path 3 Pass Gain Meter 1 Path 3 NORMAL Meter 1 ALERT Unit 1 ALERT Meter 1 Path 3 Alert --

Gain Meter 1 Path 3 ALERT Meter 1 NORMAL Unit 1 NORMAL Meter 1 Path 3 Pass Gain Meter I Path 3 NORMAL Meter 1 ALERT Unit 1 ALERT Meter 1 Path 3 Alert --

Gain Meter 1 Path 3 ALERT Meter 1 NORMAL Unit 1 NORMAL Meter 1 Path 3 Pass --

Gain Meter 1. Path 3 NORMAL Meter 1 ALERT Unit I ALERT Meter 1 Path 3 Alert --

Gain Meter 1 Path 3 ALERT Meter 1 NORMAL Unit 1 NORMAL Meter 1 Path 3 Pass --

Gain Meter 1 Path 3 NORMAL Meter 1 ALERT Unit 1 ALERT I

Unit 1 19:01:03 2001/09/07 I

Configuration Files ALARM.INI FAT.INI HYDRAULI.INI METER.INI PARAMETR. INI P CONFIG.INI PROPERTY.INI SETUP.INI 2000/12/12 2000/12/12 2001/09/07 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 18:15:40 18:15:40 17:41:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 18:15:40 FFFED282 FFFFEB2F FFFF453B FFFD66BF FFFB8AEO FFFF82DC FFFF6C54 FFFF9D29 FFF89717 FFF899D5 FFF899D5 FFF899D5 FFF899D5 FFF899D5 FFF89905 FFF899D5 I

I 3

Setup Files Setapul.txt Setapu2.txt Setapu3.txt Setapu4.txt Setapu5.txt Setapu6.txt Setapu7.txt Setapu8.txt 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 2000/12/12 Flow:

Flow:

Flow:

Flow:

Unit Unit Unit Unit 11 1

1 Current Average Maximum Minimum 81.49 81.59 82.37 80.99 0.22 Unit I Deviation Flow:

IJ JJ J

Unit I Current Temp:

Unit 1 Average Temp:

Unit 1 Maximum Temp:

Unit 1 Minimum Temp:

Unit 1 Deviation Temp:

Unit I Current System Status:

Unit 1 Minimum System Status; Unit 1 Current Mass Flow:

Unit I Average Mass Flow:

Unit 1 Maximum Mass Flow:

Unit 1 Minimum Mass Flow:

Unit 1 Deviation Mass Flow:

Unit 1 Uncertainty:

Meter 1 Current Flow:

Meter 1 Average Flow:

Meter 1 Maximum Flow:

Meter 1 Minimum Flow:

Meter 1 Deviation Flow:

Meter 1 Current Temp:

Meter 1 Average Temp:

Meter 1 Maximum Temp:

Meter 1 Minimum Temp:

Meter I Deviation Temp:

442.5 442.7 442.9 435.7 0.3 NORMAL FAIL 15290.738 15307.514 15454.595 15194.17.6 41.940 0.12 81.49 81.59 82.37 80.99 0.22 442.5 442.7 442.9 435.7 0.3 1

I I

I

Meter Meter Meter Meter Meter I

1 1.

1 Current Press:

Average Press:

Maximum Press:

Minimum Press:

Deviation Press:

I Meter 1 Current Meter Status:

Meter 1 Minimum Meter Status:

1161.97 1155.12 1170.75 200.00

• 0.33 NORMAL FAIL 15290.738

.15307.514 15454.595 15194.176 41.940 I1 Meter Meter Meter Meter Meter 11 1

1 1

Current Mass Flow:

Average Mass Flow:

Maximum Mass Flow:

Minimum Mass Flow:

Deviation Mass Flow:

1 1

Meter 1 Uncertainty:

Meter 1 Current Variance:

0.12 Path 1 Path 2 Path 3 Path 4 11232.52 15020.36 27588.72 16844.19 Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Average Vnorm:

Current Vnorm:

Maximum Vnorm:

Minimum Vnorm:

Deviation Vnorm:

Benchmark Vnorm:

Limit % Vnorm:

Average Gain:

Current Gain:

Maximum Gain:

Minimum Gain:

Deviation Gain:

Limit Gain:

Current Gain Up:

Current Gain Down:

Current TPGain Up:

Current TPGain Down:.

Average. S/N Ratio:

Current S/N Ratio:

Maximum S/N Ratio:

Minimum S/N Ratio:

Deviation S/N Ratio:

Average TDown:

Current TDown:

Maximum TDown:

Minimum TDown:

Deviation TDown:

Current TPTDown:

Average DeltaT:

Current DeltaT:

Maximum DeltaT:

Minimum DeltaT:

Deviation DeltaT:

0.8186 0.8302 0.8439 0.7865 0.009 0.8187 0.50 54.60 54.68 54.80 54.33 0.08 76.00 54.09 55.03 58.95 58.64 36.63 37.58 38.83 32.71 0.72 477614 477447 477801 477432 71 4000754 2015.9 2040.5 2082.0 1940.4 21.3 0.9972 1.0023

.1.0134 0.9806 0.005 0.9971 0.50 60.57 60.60 60.88 60.25 0.11 76.00 60.37 60.68 58.95 58.64 21.75 22.46 24.15 18.72 0.57 821752 821476 822098 821443 122 4000755 4606.8 4621.4 4677.9 4532.2 24.8 1.0523 1.0448

1. 07 68 1.0323 0.008 1.0519 0.50 68.16 68.17

68. 4.4 67.74 0.11 76.00 67.90 68.37 58.80 58.80 11.16 11.24 12.71 9.49 0.56 821689 821445 822059 821364 124 4000754 4852.1 4808.8 4974.6 4740.5 40.5 1.0098 1.0064 1.0456 0.9736 0.012 1.0109 0.50 56.02 55.97 56.21 55.82 0.08 76.00 55.35 56.44 58.64 58.80 31.41 33.03 33.55 28.04 0.61 477469 477324 477663 477289 71 4000756 2446.8 2433.8 2541.1 2353.5 30.8

!II Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1

1 1

1 1a 1

1 I

1 1

Meter 1 Current TPDeltaT:

Meter 1 Current Path Status:

Meter 1 Minimum Path Status:

Meter 1 Average Reject %:

Meter 1 Current Reject %:

Meter 1 Maximum Reject %:

Meter 1 Minimum Reject %:

Meter 1 Deviation Reject %:

Meter I Incoming Samples:

Meter 1 Number Failed Rejects:

0.6

-1.2

-0.4

-2.0 NORMAL FAIL 0.4 1.0 25.0 0.0 1.9 599 0

NORMAL FAIL.

0.5 0.8 25.5 0.0 1.8 599 0

NORMAL FAIL 8.0 8.5 31.5 0.0 2.5 599 0

NOPMAL FAIL 0.8 0.0 26.0 0.0 1.9 599.

0 1

1 1~

2 I

.1 Alarm Log Events 2001/09/07 18:11:11 2001/09/07 18:11:11 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:27 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 16:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:11:32 2001/09/07 18:21:16 Meter 1 Fail --

Unit 1 FAIL Meter 1 Path 1 Meter 1 Path 1 Meter 1 Path 2 Meter 1 Path 2 Meter 1 Path 3 Meter 1 Path 3 Meter 1 Path 4 Meter 1 Path 4 Meter 1 NORMAL Unit 1 NORMAL Meter 1 Path 1 Meter

. Path 1 Meter I Path 2 Meter I Path 2 Meter 1 Path 3 Meter 1 Path 3 Meter 1 Path 4 Meter 1 Path 4-Path Failure Fail Pass Fail Pass Fail Pass Fail Pass (APU)

Not Responding Transit Time (APU)

Not Responding

-- Transit Time (APU)

Not Responding Transit Time (APU)

-- Not Responding Transit Time Pass (APU)

NORMAL Pass (APU)

NORMAL Pass (APU)

NORMAL Pass (APU)

NORMAL Responding Responding Responding Responding 11 Verification Test Performed 1*

Watts Bar Data taken from commissioning and from plant personnel during the velocity profile alarm Unit I 8129/01 917/01

-0.861136 0.8648 0.8186

-0.339981 1.0277 0.9972 0.33998 1.0402 1.0523 0.86114 0.8996 1.0098 S/L 0.853 0.892

rO ra M(

Plant Name:

Susquehanna Unit 2 Loop A Feedwater Measurement System:

LEFM,/

Installation Geometry:

10 Diameters Downstream from a 900 Bend Non-planar bend 21 Diameters Upstream Chordal Meter Measurement Error =0.05%

P-i f R0.s ER-262 Rev. 0 Count on Caldon Appendix F

Aga a2 r

pIII TYPICAL PIPING CONFIGURATION A~'Jfl I FFV I OC~ATIrThi AND l~

It'.**""**v "F

I QA I Ql" PP&L SUSQUEHANNA LOOPA ER-262 Rev. 0 Count on Caldon Appendix F

Data Received by Plant Personnel VNORM VNORM VNORM VNORM Short TIME M1, P1 M1, P2 M1, P3 MI, P4 AvQ.

Long Avg.

SIL DATE 133 10/6/01 12:37:04 0.931569 1.037217 1.023558 0.857652 0.894611 1.030387 0.868 134 10/6/01 13:37:09 0.95617 1.041084 1.014009 0.85275 0.90446 1.027547 0.880 135 10/6/01 14:37:14 0.981016 1.04572 1.003372 0.848733 0.914874 1.024546 0.893 136 10/6/01 15:37:19 0.983483 1.046115 1.002563 0.847688 0.915586 1.024339 0.894 137 10/6/01 16:37:24 0.976356 1.043928 1.006061 0.850263 0.91331 1.024995 0.891 138 10/6/01 17:37:30.0.972266 1.043657 1.007316 0.85096 0.911613 1.025486 0.889 139 10/6/01 19:0'5:03 0.939903 1.008246 0.974203 0.823093 0.881498 0.991224 0.889 140 10/6/01 20:05:09 0.971267 1.04306 1.00795 0.851826 0.911546 1.025505 0.889 141 10/6/01 21:05:14 0.970075 1.042778 1.008554 0.851899 0.910987 1.025666 0.888 142 10/6/01 22:05:19 0.968781 1.042657 1.008944 0.852263 0.910522 1.0258 0.888 143 10/6/01 23:05:24 0.968545 1.042203 1.009547 0.85198 0.910263 1.025875 0.887 144 10/7/01 0:05:29 0.968619 1.042056 1.009591 0.852257 0.910438 1.025824 0.888 145 10/7/01 1:05:35 0.967196 1.041938 1.010146 0.85217 0.909683 1.026042 0.887 146 10/7/01 2:05:40 0.966325 1.041626 1.010619 0.852474 0.9094 1.026123 0.886 147 10/7/01 3:05:45 0.966818 1.042383 1.009713 0.852497 0.909657 1.026048 0.887 148 10/7/01 4:05:50 0.967062 1.041676 1.010334 0.852551 0.909806 1.026005 0.887 149 10/7/01 5:05:55 0.963437 1.041647 1.0112881 0.852982 0.908209 1.026468 0.885

0

~

Plant Name:

Susquehanna Unit 2 Loop B Feedwater Measurement System:

LEFM-/

Installation Geometry:

10 Diameters Downstream from a 90' Bend Non-planar bend 17 Diameters Upstream

-.5

~

0 P-1 df R~dln ER-262 Rev. 0 Count on Caldon Appendix F

R WE TYPICAL PIPING CONFIGURATION AND LEFM LOCATION PP&L SUSQUEHANNA 0

LJA.13 LOOP B CH.

SKETCH SKRSH-35B.DWG ER-262 Rev. 0 Count on Caldon Appendix F

Data Received by Plant Personnel VNORM VNORM VNORM VNORM Short TIME M2, P1 M2, P2 M2, P3 M2, P4 Avg.

Long Avg.

S/L DATE 95 1014101 22:34:47 0.8584 1.029299 1.046331 0.87931 0.868855 1.037815 0.837 96 10/4/01 23:34:52 0.858327 1.029064 1.046604 0.879243 0.868785 1.037834 0.837 97 10/5/01 0:34:57 0.858075 1.029354 1.046389 0.879237 0.868656 1.037872 0.837 98 10/5/01 1:35:02 0.858352 1.029276 1.046372 0.879282 0.868817 1.037824 0.837 99 10/5/01 2:35:07 0.85833 1.029248 1.046464 0.879096 0.868713 1.037856, 0.837 100 10/5/01 3:35:13 0.857994 1.02937 1.046337 0.879438 0.868716 1.0378541 0.837 101 10/5/01 4:35:18 0.858446 1.029274 1.046358 0.879252 0.868849 1.037816 0.837 102 10/5/01 5:35:23. 0.858421 1.029451 1.046229 0.879113 0.868767 1.03784 0.837 103 10/5/01 6:35:28 0.858379 1.029293 1.046403 0.879103 0.868741 1.037848 0.837 104 10/5/01 7:35:33 0.858999 1.029274 1.046164 0.879361 0.86918 1.037719 0.838 105 10/5/01 8:35:38 0.858118 1.029351 1.046412 0.879134 0.868626 1.037881 0.837 106 10/5/01 9:34:44 0.857948 1.029428 1.046339 0.879293 0.868621 1.037883 0.837 107 10/5/01 10:34:49 0.858131 1.029239 1.046535 0.87908 0.868606 1.037887 0.837 108 10/5/01 11:34:54 0.858522 1.029101 1.046441 0.879489 0.869005 1.037771 0.837 109 10/5/01 12:34:59 0.85829 1.029324 1.04635 0.879256 0.868773 1.037837 0.837 110 10)5/01 13:35:04 0.858456 1.029479 1.046244 0.878926 0.868691 1.037861 0.837

17==

0 r--Nr---v-N 0 Plant Name:

Susquehanna Unit 2 Loop C Feedwater Measurement System:

LEFM,/

Installation Geometry:

10 Diameters Downstream from a 90' Bend Non-planar bend 17 Diameters Upstream Chordal Meter Measurement Error = 0.01%

.0.5 0

0.

P-df R~dh.

ER-262 Rev. 0 Count on Caldon Appendix F

0 rz:~~ 0 4

60ggý SPI cL SKETCH SKRSH-35C,DWG TYPICAL PIPING CONFIGURATION AND LEFIM LOCATION 1.

. 1.

I PP&L SUSQUEHANNA LOOP Q ER-262 Rev. 0 Count on Caldon Appendix F

Data Received by Plant Personnel VNORM VNORM VNORM VNORM Short TIME M3, P1 M3, P2 M3, P3 M3, P4 Avg.

Long Avg.

SIL DATE 109 10/5/01 12:34:59 0.870465 1.030248 1.0488 0.855393 0.862929 1.039524 0.830 110 10/5/01 13:35:04 0.869863 1.030255 1.048903 0.855594 0.862729 1.039579 0.830 111 10/5/01 14:35:10 0.869964 1.030384 1.048817 0.855356 0.86266 1.039601 0.830 112 10/5101 15:35:15 0.870409 1.030193 1.048877 0.855363 0.862886 1.039535 0.830 113 10/5/01 16:35:20 0.869652 1.030225 1.049009 0.855552 0.862602 1.039617 0.830 114 10/5/01 17:35:25 0.86979 1.030176 1.048998 0.855629 0.86271 1.039587 0.830 115 1015/01 18:35:30 0.869946 1.03033 1.048979 0.855002 0.862474 1.039&54 0.830 116 10/5/01 19:35:35 0.870376 1.030603 1.048459 0.855435 0.862905 1.039531 0.830 117 10/5/01 20:35:41 0.869924 1.030366 1.048768 0.855638 0.862781 1.039567 0.830 118 10/5/01 21:35:46 0.870015 1.030551 1.048605 0.855456 0.862735 1.039578 0.830 119 10/5/01 22:35:51 0.870349 1.03016 1.049047 0.854947 0.862648 1.039603 0.830 120 10/5/01 23:35:56 0.87075 1.030298 1.048586 0.855666 0.863208 1.039442 0.830 121 10/6/01 0:36:01 0.870223 1.030536 1.0486 0.855323 0.862773 1.039568 0.830 122 10/6/01 1:36:07 0.869851 1.030667 1.048538 0.855451 0.862651 1.039603 0.830 123 10/6/01 2:36:12 0.869714 1.030353 1.048966 0.855188 0.862451 1.039659 0.830 124 10/6/011 3:36:17 0.870174 1.030264 1.048833 0.85551 0.862842 1.039548 0.830 125 10(6/01 4:36:22 0.870365 1.030263 1.048813 0.855386 0.862876 1.039538 0.830

Susquehanna Unit 2 12:09:22 2001/05/04 Configuration Files ALARM.INI FAT.INI HYDRAULI.INI METER.INI PARAMETR.IN!

P CONFIG.INI PROPERTY.INI SETUP.INI Setup Files Setapul.txt Setapu2.txt Setapu3.txt Setapu4.txt 2001/05/04 2001/04/16 2001/05/04 2001/05/03 2001/04/24 2.001/05/03 2001/04/16 2001/05/04 2001/05/03 2001/05/03 2001/05/03 2001/04/16 11:46: 46 20:54:32 11:45:52 16:47:26 15:06:08 16:01:44 21: 17: 40 11:41:54 08:40:10 10:13:30 08:40:48 21:4 6:14 FFFF5F6D FFFFD4A7 FFFF9407 FFFD2091 FFFC6D3D FFFEA975 FFFFEC75 FFFEE167 FFFE17FD FFFE17FD FFFE17F7 FFFE18E7 71.18 71.17 71.24 71.06 0.04 Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Unit Unit Unit Unit Unit Unit Unit Unit Unit Unit 2

2 2

2 2

Current Average Maximum Minimum Flow:

Flow:

Flow:

Flow:

Deviation Flow:

2 Current Temp:

2 Average Temp:

2 Maximum Temp:

2 Minimum Temp:

2 Deviation Temp:

Susquehanna Unit 2 Current System Status:

Susquehanna Unit 2 Minimum System Status:

385.8 385.7 385.8 370.0 1.0 NORMAL FAIL 13.970

13. 968 14.111 13.950 0.012 Susquehanna Susquehanna Susquehanna Susquehanna Susquehanna Unit Unit Unit Unit Unit 2

2 2

2 2

Current Mass Flow:

Average Mass Flow:

Maximum Mass Flow:

Minimum Mass Flow:

Deviation Mass Flow:

Susquehanna Unit 2 Uncertainty:

0.03 Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1 Current Flow:

1 Average Flow:

1 Maximum Flow:

1 Minimum Flow:

1 Deviation Flow:

2 2

2 2

2 3

3 3

3 Current Flow; Average Flow:

Maximum Flow:

Minimum.Flow:

Deviation Flow:

Current Flow:

Average Flow:

Maximum Flow:

Minimum Flow:

23.66 23.68 23.77 23.61 0.04 23.81 23.84 23.94 23.75 0.05 23.71 23.65 23.82 23.47

Meter 3 Deviation Flow:

00 0.09 Meter 1 Current Tempt 387.1 Meter 1 Average Tempt 387.0 Meter 1 Maximum Temp:

387.1 Meter 1 Minimum Tempt 311.3 Meter 1 Deviation Temp:

1.0 Meter 2 Current Temp:

385.4 Meter 2 Average Tempt 385.4 Meter 2 Maximum Tempt 385.5 Meter 2 Minimum Tempt 369.6 Meter 2 Deviation Tempt 1.0 Meter 3 Current Tempt 384.8 Meter 3 Average Tempt 384.8 Meter 3 Maximum Tempt 384.9 Meter 3 Minimum Tempt 369.0 Meter 3 Deviation Tempt 1.0 Meter 1 Current Press:

1105.00 Meter 1 Average Press:

1002.21 Meter 1 Maximum Press:

1105.00 Meter 1 Minimum Press:

0.00 Meter 1 Deviation Press:

1.03 Meter 2 Current Press:

1106.10 Meter 2 Average Press:

1003.21 Meter 2 Maximum Press:

1106.10 Meter 2 Minimum Press:

0.00 Meter 2 Deviation Press:

1.03 Meter 3 Current Press:

1104.40 Meter 3 Average Press:

1001.67 Meter 3 Maximum Press:

1104.40 Meter 3 Minimum Press:

0.00 Meter 3 Deviation Press:

1.03 Meter 1 Current Meter Status:

NORMAL Meter 1 Minimum Meter Status:

FAIL Meter 2 Current Meter Status:

NORMAL Meter 2 Minimum Meter Status:

FAIL Meter 3 Current Meter Status:

NORMAL Meter 3 Minimum Meter Status:

FAIL Meter 1 Current Mass Flow:

4.639 Meter 1 Average Mass Flow:

4.643 Meter 1 Maximum Mass Flow:

4.693 Meter 1 Minimum Mass Flow:

4.629 Meter 1 Deviation Mass Flow:

0.009 Meter 2 Current Mass Flow:

4.615 Meter 2 Average Mass Flaw:

4.679 Meter 2 Maximum Mass Flow:

4.729 Meter 2 Minimum Mass Flow:

4.662

Meter 2 Deviation Mass Flow:

Meter 3 Current Mass Flow:

Meter 3 Average Mass Flow:

Meter 3 Maximum Mass Flow:

Meter 3 Minimum Mass Flow:

Meter 3 Deviation Mass Flow:

Meter 1 Uncertainty:

Meter 2 Uncertainty:

Meter 3 Uncertainty:

Meter 1 Current Variance:

Meter 2 Current Variance:

Meter 3 Current Variance:

Meter 1 Average Vnorm:

Meter 1 Current Vnorm:

Meter 1 Maximum Vnorm:

Meter I Minimum Vnorm:

Meter 1 Deviation Vnorm:

Meter 1 Benchmark Vnorm:

Meter 1 Limit % Vnorm:

Meter 2 Average Vnorm:

Meter 2 Current Vnorm:

Meter 2 Maximum Vnorm:

Meter 2 Minimum Vnorm:

Meter 2 Deviation Vnorm:

Meter 2 Benchmark Vnorm:

Meter 2 Limit % Vnorm:

Meter 3 Average Vnorm:

Meter 3 Current Vnorm:

Meter 3 Maximum Vnorm:

Meter 3 Minimum Vnorm:

Meter 3 Deviation Vnorm:

Meter 3 Benchmark Vnorm:

Meter 3 Limit % Vnorm:

Meter I Average Gain:

Meter I Current Gain:

Meter 1 Maximum Gain:

Meter 1 Minimum Gain:

Meter 1 Deviation Gain:

Meter I Limit Gain:

Meter 1 Current Gain Up:

Meter 1 Current Gain Down:

Meter 1 Current TPGain Up:

Meter 1 Current TPGain Down:

Meter 2 Average Gain:

0.010 4.656 4.645 4.689

4. 609 0.018 0.06 0.04 0.04 Path 1 Path 2 Path 3 Path 4 10611.80 9480.00 6556.76 2452.37 2306.92 3502.04 3445.44 2121.49 2339.39 3411.16 3430.44 2677.44 0.9309 0.9300 0.9395 0.9220 0.003 0.9301 0.50 0.8524 0.8543 0.8551 0.8503 0.001 0.8522 0.50 0.8616 0.8602 0.8651 0.8580 0.002 0.8616 0.50 46.85 46.89 46.94 46.77 0.04 76.00 45.85 47.79 64.13 63.82 1.0403 1.0380 1.0425 1.0380 0.001 1.0399 0.50 1.0315 1.0315 1.0328 1.0300 0.001 1.0316 0.50 1.0324 1.0330 1.0342 1.0311 0.001 1.0326 0.50 50.73 50.72 50.79 50.67 0.03 76.00 50.70 50.55 63.82 63.82 1.0224 1.0243 1.0251 1.0197 0.001 1.0229 0.50 1.0490 1.0479 1.0500 1.0477 0.000 1.0490 0.50 1.0507 1.0505 1.0520 1.0495 0.001 1.0506 0.50 51.38 51.34 51.46 51.32 0.03 76.00 51.49 51.02 63.97 63.82 0.8519 0.8541 0.8547 0.8487 0.002 0.8520 0.50 0.8685 0.8703 0.8712 0.8663 0.001 0.8684 0.50 0.8500 0.8504 0.8533 0.8468 0.002 0.8500 0.50 46.50 46.61 46.62 46.43 0.03 76.00 45.69 47.48 64.13 63.97 44.93 48.41 47.81 48.25

Meter 2 Current Gain:

44.93 48.43 47.79 48.29 Meter 2 Maximum Gain; 44.98 48.46 47.86 48.32 Meter 2 Minimum Gain:

44.88 48.37 47.77 48.19 Meter 2 Deviation Gain:

0.02 0.02 0.02 0.03 Meter 2 Limit Gain:

76.00 76.00 76.00 76.00 Meter 2 Current Gain Up:

44.28 48.58 48.10 47.95 Meter 2 Current Gain Down:

45.38 48.10 47.32 48.42 Meter 2 Current TPGain Up:

63.97 63.97 63.82 63.97 Meter 2 Current TPGain Down:

63.82 63.82 63.66 63.66 Meter 3 Average Gain:

44.20 48.55 47.08 43.29 Meter 3 Current Gain:

44.20 48.56 47.09 43.23 Meter 3 Maximum Gain:

44.28 48.63 47.16 43.40 Meter 3 Minimum Gain:

44.08 48*46 46.93 43.21 Meter 3 Deviation Gain:

0.04 0.03 0.06 0.05 Meter 3 Limit Gain:

76.00 76.00 76.00 76.00 Meter 3 Current Gain Up:

43.50 48.73 47.48 42.87 Meter 3 Current Gain Down:

44.91 48.26 46.54 43.50 Meter 3 Current TPGain Up:

63.66 63.82 63.50 63.66 Meter 3 Current TPGain Down:

63.66 63.66 63.50 63.82 Meter 1 Average S/N Ratio:

97.20 97.09 96.51 96.22 Meter 1 Current S/N Ratio:

97.52 97.26 96.48 95.99 Meter I Maximum S/N Ratio:

97.70 97.38 96.84 96.75 Meter 1 Minimum S/N Ratio:

95.13 94.84 94.47 94.85 Meter 1 Deviation S/N Ratio:

0.33 0.32 0.30 0.28 Meter 2 Average S/N Ratio:

87.80 90.28 88.97 86.87 Meter 2 Current S/N Ratio:

87.98 88.75 87.07 86.46 Meter 2 Maximum S/N Ratio:

92.00 95.06 93.49 92.23 Meter 2 Minimum S/N Ratio:

84.45 88.25 86.37 84.25 Meter 2 Deviation S/N Ratio:

1.44 1.48 1.30 1.41 Meter 3 Average S/N Ratio:

18.97 56.05 41.71 15.81 Meter 3 Current S/N Ratio:

19.28 57.13 43.11 15.80 Meter 3 Maximum S/N Ratio:

19.69 57.34 43.59 16.24 Meter 3 Minimum S/N Ratio:

18.18 53.56 39.98 15.39 Meter 3 Deviation S/N Ratio:

0.30 0.63 0.88 0.16 Meter 1 Average TDown:

244697 395163 395068 244698 Meter 1 Current TDown:

244696 395164 395064 244696 Meter 1 Maximum TDown:

244713 395189 395094 244713 Meter 1 Minimum TDown:

244683 395142 395049 244687 Meter 1 Deviation TDown:

6 10 10 5

Meter 1 Current TPTDown:

4500555 4500554 4500554 4500556 Meter 2 Average TDown:

244412 394581 394223 244057 Meter 2 Current TDown:

244411 394583 394227 244057 Meter 2 Maximum TDown:

244427 394605 394247 244071 Meter 2 Minimum TDown:

244398 394556 394201 244044 Meter 2 Deviation TDown:

7 11 11 6

Meter 2 Current TPTDown:

4500594 4500596 4500599 4500595 Meter 3 Average TDown:

243956 393939 394092 243980 Meter 3 Current TDown:

243952 393929 394083 243975 Meter 3 Maximum TDown:

243967 393962 394113 243993 Meter 3 Minimum T~own:

243944 393920 394074 243968

Meter 3 Deviation TDown:

7 12 11 7

Meter 3 Current TPTDown:

4500464 4500468 4500464 4500462 Meter 1 Average DeltaT:

1135.4 2346.1 2301.8 1043.4 Meter 1 Current DeltaT:

1133.2 2339.3 2310.0 1045.0 Meter I Maximum DeltaT:

1146.3 2355.9 2321.6 1049.8 Meter 1 Minimum DeltaT:

1126.9 2338.6 2299.8 1039.2 Meter 1 Deviation DeltaT:

4.3 5.0 4.7 2.7 Meter 1 Current TPDeltaT:

-0.6 2.2 2.2

-0.6 Meter 2 Average DeltaT:

1043.2 2315.6 2352.7 1047.8 Meter 2 Current DeltaT:

1044.5 2313.6 2348.0 1048.9 Meter 2 Maximum DeltaT:

1048.5 2324.6 2363.7 1053.5 Meter 2 Minimum DeltaT:

1038.3 2308.0 2343.4 1043.6 Meter 2 Deviation DeltaT:

2.2 4.4 5.2 2.5 Meter 2 Current TPDeltaT:

-2.9

-0.9

-3.8

-1.1 Meter 3 Average DeltaT:

1041.0 2306.7 2349.4 1031;1 Meter 3 Current DeltaT:

1041.8 2313.5 2354.8 1034.1 Meter 3 Maximum DeltaT:

1052.6 2322.7 2367.0 1039.6 Meter 3 Minimum DeltaT:

1034.5 2287.3 2332.7 1021.2 Meter 3 Deviation DeltaT:

4.2 9.4 8.7 4.9 Meter 3 Current TPDeltaT:

0.4

-4.2 2.2

-2.2 Meter 1 Current Path Status:

NORMAL NORMAL NORMAL NORMAL Meter I Minimum Path Status:

FAIL FAIL FAIL FAIL Meter 2 Current Path Status:

NORMAL NORMAL NORMAL NORMAL Meter 2 Minimum Path Status:

FAIL FAIL FAIL FAIL Meter 3 Current Path Status:

NORMAL NORMAL NORMAL NORMAL Meter 3 Minimum Path Status:

FAIL FAIL FAIL FAIL Meter I Average Reject %:

0.1 0.1 0.1 0.1 Meter 1 Current Reject 1:

0.0 0.0 0.0

.0.0 Meter 1 Maximum Reject %:

4.2 4.2 4.2 4.2 Meter 1 Minimum Reject %:

0.0 0.0 0.0 0.0 Meter 1 Deviation Reject %:

0.5 0.5 0.5 0.5 Meter 1 Incoming Samples:

258 258 258 258 Meter-I Number Failed Rejects:

0 0

0 0

Meter 2 Average Reject %:

0.0 0.0 0.0 0.0 Meter 2 Current Reject %:

0.0 0.0 0.0 0.0 Meter 2 Maximum Reject %:

0.0 0.0 0.0 0.0 Meter 2 Minimum Reject %:

0.0 0.0 0.0 0.0 Meter 2 Deviation Reject %:

0.0 0.0 0.0 0.0 Meter 2 Incoming Samples:

258 258 258 258 Meter 2 Number Failed Rejects:

0 0

0 0

Meter 3 Average Reject 1:

0.0 0.0 0.0 0.1 Meter 3 Current Reject %:

0.0 0.0 0.0 0.0 Meter 3 Maximum Reject 1:

0.0 0.0 0.0 0.3 Meter 3 Minimum Reject %:

0.0 0.0 0.0 0.0 Meter 3 Deviation Reject %:

0.0 0.0 0.0 0.1 Meter 3 Incoming Samples:

258 258 258 258 Meter 3 Number Failed Rejects:

0 0

0 0

Hydrauli.ini DEPFAULTCFRATIOI

, 1. 0000,0.9999,1.0003, 0.9998 DEFAULTCFRATIO2 :1.

0003,1.0000, 0.9999°,0.9999 DEFAULTCERATIO3 :1.0002.1..00.i.00.O00..9999 DEFAULTVELOCLTYI :,0.9328,1.0400,1.0-217,.O8531 DEFAULTVELOCITY2:, 0.8533, 1.0311, 1.0489,0. 869a DEFAULTVELOCITY3 :, O. 8632, 1.0323,1. 0502,'0. 8507 SOUNDVELOCrTYNOMI :, 50300 SOUNDVELVICITYNOM2:, 50300 SOUNfDVELCCITYNOM3 :, 5300 PROFILEFACTORCOEFA0L:,1. 0038E÷000 PROFILEFACTORCOEFAP02:, L. 0101E+000 PROFILEFACTORCOEFA03 :,1. 0068E+000 KAXIN :,720 Page I

PP&L Unit 2 Data taken from commissioning and from plant personnel during the velocity profile alarm Meter 1 5/4/01 10/6/01

-0.861136 0.9310 0.9510

-0.339981 1.0403 1.0391 0.33998 1.0223 1.0167 0.86114 0.8519 0.8555 S/L 0.864 0.879 Meter 2 5/4/01 10/5/01

-0.861136 0.8524 0.8581

-0.339981 1.0315 1.0295 0.33998 1.0490 1.0463 0.86114 0.8685 0.8793 S/L 0.827 0.837 Meter 3 5/4/01 10/5101

-0.861136 0.8617 0.8703

-0.339981 1.0324 1.0304 0.33998 1.0507 1.0488 0.86114 0.8500 0.8551 S/L 0.822 0.830

0

-7n.0 Plant Name:

Indian Point Unit 2 Loop 21 Feedwater Measurement System:

LEFM,/

Installation Geometry:

10 Diameters Downstream from a 900 Elbow Non-planar bend 10 Diameters Upstream Chordal Meter Measurement Error= 0.02%

-I 0.30 0.3 ER-262 Rev. 0 Count on Caldon Appendix F

Plant Name:

Feedwater Measurement System:

Installation Geometry:

Indian Point Unit 2 Loop 22 LEFM,/

12 Diameters Downstream from a 90* Elbow Non-planar bend 10 Diameters Upstream

-1 0

rFm*s 4 1 a~

0.5 ER-262 Rev. 0 Count on Caldon Appendix F

M0 0

Plant Name:

Feedwater Measurement System:

Installation Geometry:

Indian Point Unit 2 Loop 23 LEFM,/

15 Diameters Downstream from a 90* Elbow Non-planar bend 10 Diameters Upstream

-I 0.50 0.5 Fament of Radl..

ER-262 Rev. 0 Count on Caldon Appendix F

Plant Name:

Indian Point Unit 2 Loop 24 Feedwater Measurement System:

LEFM,/

Installation Geometry:

13 Diameters Downstream from a 900 Elbow Non-planar bend 10 Diameters Upstream 0

0.5 Fmg d

.R~dim ER-262 Rev. 0 Count on Caldon Appendix F

r zD 1 0 TYPICAL PIPING CONFIGURATION ER-262 Rev. 0 Count on Caldon Appendix F

Indian Point 2 Data taken from trip reports and commissioning data Loop 21 7/8/95 1/12/96 9/24/98 10/25/99 10/14/01

-0.861136 0.9834 1.0372 0.9099 0.8412 0.8763

-0.339981 1.0534 1.0837 1.0229 0.9868 1.0070 0.33998 0.9937 0.9671 1.0323 1.0684 1.0503 0.86114 0.8445 0.7954 0.9077 0.9762 0.9468 S/L 0.893 0.894 0.884 0.884 0.886 Loop 22 7/8/95 1/12/96 9/24/98 10/25/99 10/14/01

-0.861136 0.8920 0.8805 0.8943 0.8822 0.8744

-0.339981 0.9978 0.9933 1.0065 1.0053 1.0144 0.33998 1.0315 1.0535 1.0359 1.0460 1.0411 0.86114 0.9974 0.9661 0.9675 0.9489 0.9411 S/L 0.931 0.902 0.912 0.893 0.883 Loop 23 7/8/95 1/12/96 9/24/98 10/25/99 10/14/01

-0.861136 0.8783 0.8845 0.8122 0.7453 0.7813

-0.339981 1.0019 1.0058 0.9925 0.9696 0.9711 0.33998 1.0345 1.0379 1.0496 1.0907 1.0847 0.86114 0.9865 0.9727 1.0508 1.0543 1.0328 S/L 0.916 0.909 0.912 0.873 0.882 Loop 24 7/8/95 1/12/96 9/24/98 10/25/99 10/14/01

-0.861136 0.8257 0.8087 0.8679 0.8840 0.8822

-0.339981 0.9733 0.9726 0.9801 0.9972 1.0042 0.33998 1.0594 1.0675 1.0498 1.0390 1.0285 0.86114 1.0520 1.0611 1.0375 0.9997 0.9964 S/L 0.924 0.917 0.939 0.925 0.924

13

,ýE Plant Name:

Feedwater Measurement System:

Installation Geometry:

Indian Point Unit 3 Loop 31 LEFM,/

5.8 Diameters Downstream from a 90' Elbow Non-planar bend 10 Diameters Upstream Chordal Meter Measurement Error= 0.03%

(.1*

zI F~Ti~1

-I

-0.5 0

P-atn f JiRdl-0.5 ER-262 Rev. 0 Count on Caldon Appendix F

=

.z g-M Plant Name:

Indian Point Unit 3 Loop 32 Feedwater Measurement System:

LEFMI Installation Geometry:

5.8 Diameters Downstream from a 900 Elbow Non-planar bend 10 Diameters Upstream

.2 030 0.5 ER-262 Rev. 0 Count on Caldon Appendix F

!ý Q0 Plant Name:

Indian Point Unit 3 Loop 33 Feedwater Measurement System:

LEFMvr Installation Geometry:

5.8 Diameters Downstream from a 900 Elbow Non-planar bend 10 Diameters Upstream Chordal Meterj Measurement Error < 0.04%

-0.5 0

0.5 ER-262 Rev. 0 Count on Caldon Appendix F

0 ~

0 J~L~

Plant Name:

Indian Point Unit 3 Loop 34 Feedwater Measurement System:

LEFMI" Installation Geometry:

5.8 Diameters Downstream from a 90' Elbow Non-planar bend 10 Diameters Upstream Chordal Meter Measurement Error = 0.03%

ft1 0.5~

ER-262 Rev. 0 Count on Caldon Appendix F

£2 ~

0 TYPICAL PIPING CONFIGURATION SKECHc-SKRSH-2g.DWG ER-262 Rev. 0 Count on Caldon Appendix F

Indian Point 3 Data from remote monitoring program - found under LEFMLOGS Loop 31 6/23/98 8/26/99 11/3/99 6/23/00 12/10/00 6/21/01 10/13/01

-0.861136 0.906 0.898 0.942 0.895 0.898 0.894 0.891

-0.339981 0.999 0.991 0.990 0.995 0.988 1.003 0.999 0.33998 1.034 1.037 1.034 1.032 1.039 1.030 1.034 0.86114 0.966 0.990 0.960 0.990 0.993 0.979 0.981 S/L 0.921 0.931 0.940 0.930 0.933 0.921 0.921 Loop 32 6/23/98 8/26/99 11/3/99 6/23/00 12/10/00 6/21/01 10/13/01

-0.861136 0.846 0.849 0.845 0.838 0.847 0.845 0.851

-0.339981 0.978 0.979 0.983 0.976 0.978 0.980 0.982 0.33998 1.057 1.054 1.050 1.058 1.055 1.053 1.049 0.86114 1.018 1.028 1.028 1.031 1.023 1.025 1.028 S/L 0.916 0.923 0.921 0.919 0.920 0.920 0.925 Loop 33 6/23/98 8/26199 1113/99 6/23/00 12/10/00 6/21/01 10/13/01

-0.861136 0.992 0.982 1.024 0.996 0.981 0.968 1.000

-0.339981 1.028 1.030 1.049 1.018 1.012 1.019 1.036 0.33998 0.998 0.996 0.979 1.000 1.009 1.006 0.991 0.86114 0.902 0.907 0.868 0.925 0.931 0.932 0.891 S/L 0.935 0.932 0.933 0.952 0.946 0.938 0.933 Loop 34 6/23/98 8/26/99 11/3/99 6/23/00 12/10/00 6/21/01 10/13/01

-0.861136 0.996 0.961 0.953 0.964 0.950 0.967 0.956

-0.339981 1.000 0.993 0.995 1.000 0.994 1.002 1.001 0.33998 1.008 1.019 1.018 1.019 1.020 1.013 1.015 0.86114 0.963 0.984 0.991 0.957 0.988 0.969 0.974 S/L 0.976 0.967 0.966 0.951 0.962 0.961 0.957

i ak 0

Plant Name:

Feedwater Measurement System:

Installation Geometry:

Comanche Peak Unit 1 LEFM,/

11.2 Diameters Downstream of a 90* Elbow Non-planar feeds 18 Diameters Upstream Chordal Meter Measurement Error < 0.01%

(... 7

.0.5 0

0.3 ER-262 Rev. 0 Count on Caldon Appendix F

SKETCH SKRSH-3O.DWG Uýt 30 TYPICAL PIPING CONFIGURAllON AND LEFM LOCAT1ON COMANCHE PEAK 1 ER-262 Rev. 0 Count on Caldon Appendix F

1-rI0 r-e-r 03 Plant Name:

Feedwater Measurement System:

Installation Geometry:

Comanche Peak Unit 2 LEFMI 11.2 Diameters Downstream of a 90' Elbow Non-planar feeds 18 Diameters Upstream Chordal Meter Measurement Error<0.01%

.0.3 0

P-t

.( 1 Rdk.dho 0.5 ER-262 Rev. 0 Count on Caldon Appendix F

SKETCH SKRSH-31.DWG TYPICAL PIPING CONFIGURATION AND LEFM LOCATION COMANCHE PEAK 2 ER-262 Rev. 0 Count on Caldon Appendix F

Comanche Peak Data taken from commissioning and from plant personnel Unit 1 11/3/99 3/31/00

-0.861136 1.0071 1.0069

-0.339981 1.0515 1.0513 0.33998 0.9858 0.9882 0.86114 0.8635 0.8565 S/L 0.918 0.914 Unit 2 10/10/99 3/31/00

-0.861136 0.9262 0.9265

-0.339981 1.0177 1.0173 0.33998 1.0237 1.0245 0.86114 0.9304 0.9283 S/L 0.909 0.908

Plant Name:

Feedwater Measurement System:

Installation Geometry:

Prairie Island Unit 2 Loop A LEFMI/

20 Diameters Downstream from a 90* Bend Non-planar bend 4 Diameters Upstream Chordal Meter Measurement Error = 0.03%

-I

-. 5a 0.5 F-0u dE PRadh ER-262 Rev. 0 Count on Caldon Appendix F

a U

5I~~w~U Plant Name:

Prairie Island Unit 2 Loop B Feedwater Measurement System:

LEFMV Installation Geometry:

20 Diameters Downstream from a 900 Bend Non-planar bend 4 Diameters Upstream Chordal Meter Measurement Error = 0.02%

-I 0.50 0.5 P....t af Rai~um ER-262 Rev. 0 Count on Caldon Appendix F

==Lj a cz,=0 ff jr-=egma SKETCH-SKRSH-32.DWG So,.

t I

3-:

30-I TYPICAL PIPING CONFIQHRAT1ON T*"

w w-AND LEFM LOCATION PRAII[

ILANDl ER-262 Rev. 0 Count on Caldon Appendix F

5/17198 0.870755 1.015193 1.043678 0.925072 0.897914 1.029436 0.872 5/22/98 0.871222 1.014476 1.042591 0.930922 0.901072 1.028534 0.876 5/26/98 0.871755 1.015432 1.042008 0.929041. 0.900398 1.02872 0.875 5/29/98. 0,870902 1.015041 1.042119 0.930853 0.900877 1.02858 0.876 6/1/98 0.87008 1.014307 1.043693 0.926702 0.899391 1.029 0.874 614/98 0.872683 1.016676 1.039385 0.932975 0.902829 1.028031 0.878 616/98 0.869605 1.014675 1.042761 0.931202 0.900404 1.028718 0.875 6/9/98 0.872216 1.015321 1.043119 0.925113 0.898665 1.02922 0.873

.6/12/98 0.870692 1.014145 1.043359 0.929805 0.900249 1.028752 0.875 6115198 0.871853 1.015781 1,041277 0.930322 0.901087 1.028529 0.876 6/18/98 0.8715 1.015434 1.043109 0,925388 0.898444 1.029272 0.873 6/24/98 0.873052 1.016014 1.042197 0.925216 0.899134 1.029106 0.874 6/27/98 0.871426 1.012633 1.043414 0.934219 0.902822 1.028024 0.878 6130/98 0.870716 1.016565 1.04245 0.924561 0,897638 1.029508 0,872 7/3198 0,869836 1.014878 1.042936 0.929472 0.899654 1.028907 0.874 7/7198 0.870063 1.014531 1.042797 0.931189 0.900626 1.028664 0.876 7/10/98 0.872366 1.016237 1.042103 0.925272 0.898819 1.02917 0,873 7/14/98 0.8711 1.015548 1.041975 0;929519 0.90031 1.028762 0.875 7/17198 0.871768 1.016207 1.043938 0.919757 0.895763 1.030073 0.870 7/18/98 0.87218 1.015789 1.042993 0.923961 0.89807 1.029391 0,872 7/20/98 0.870761 1.014915 1.043325 0.927327 0.899044 1.02912 0.874 7/23198 0.870269 1.014806 1.04286 0.929668 0.899969 1.028833 0.875 7130/98 0.870636 1.01571 1.042882 0.926208- 0.898422 1.029296 0.873 8/2/98 0.871142 1.014613 1.042983 0.928955 0.900049 1.028798 0.875 8/4198 0.87158 1.015787 1.0444 0.919716 0.895648 1.030094 0.869 8/7/98 0.871319 1.01559 1.043088 0.92525 0.898285 1.029339 0.873 8/7/98 0.871319 1.01559 1.043088 0.92525 0.898285 1.029339 0.873 8/10/98 0.871261 1.015809 1.042036 0.928128 0.899695 1.028923 0.874 8/13/98 0.871016 1.014967 1.042444 0.929883

-0.90045 1.028706 0.875 8/16198 0.87071 1.015656 1.0437 0.923476 0.897093 1.029678 0.871 8/19/98 0.871419 1.0159 1.042314 0.926708 0.899064 1.029107 0.874 8/25/98 0.871231 1.01602 1.043287 0.923017 0.897124 1.029654 0.871 8/27/98 0.872619 1.015624 1.041902 0.927836 0.900227 1.028763 0.875 8/30/98 0.872191 1.017478 1.039844 0.929028 0,90061 1.028661 0.876 9/2/98 0.87168 1.015657 1.043692 0.922428 0.897054 1.029675 0.871 9/5/98 0,870872 1.014593 1.043703 0.92696 0.898916 1.029148 0.873 918/98 0.870322 1.015253 1.042978 0.927705 0.899013 1.029115 0.874 9/12/98 0.869388 1.016051 1.044317 0.921317 0.895353 1.030184 0.869 9/15/98 0.870605 1.015736 1.043186 0.925086 0.897846 1.029461 0.872 9/18/98 0.86963 1.015043 1.043667 0.926674 0.898152 1.029355 0.873 0.879767 1.018073 1.036881 0.929628 0.904697 1.027477 0.881 9/24/98 0.872419 1.017812 1.0422 0.919366 0.895893 1.030006 0.870 9/27/98 0.869978 1.015507 1.043576 0.925059 0,697518 1.029542 0.872 9/30/98 0.870497 1.015837 1.043162 0.924942 0.897719 1.0295 0.872 10/3/98 0.87028 1.015459 1.043248 0.92613 0,898205. 1.029354 0.873 0.870836 1.016599 1.046326 0.910981 0.890908 1.031463 64 10/9/98 0.870322 1.015834 1.044506 0.920366 0.895344 1.03017 0.869 10/12/98 0.871669 1.015933 1.043209 0.923216 0.897443.1.029571 0.872

.10/15/98 0.872225 1.016998 1.043822 0.91683 0.894528 1.03041 0.868

10/18/98 0.873035 1.017446 1.042911 0.917664 0.89535 1.030179 0.869 10/21/98 0.870439 1.015696 1.043728 0.923475 0.896957 1.023712 0.871 10/24/98 0.870113 1.015195 1.043844 0.925127 0.89762 1.02952 0.872 10127/98 0.870246 1.015832 1.043589 0.923603 0.896925 1.029711 0.871 10/30/98 0.87015 1.015226 1.043695 0.925691 0.897921 1.029461 0.872 11/2/98 0.870786 1.015424 1.042805 0.927208 0.898997 1.029115 0,874.

1115198 0.871564 1.015532 1.042504 0.927166 0.899365 1.029018 0.874 11/8/98 0.871828 1.017437 1.044747 0.912569 0.892199 1.031092 0.865 Loop 31 M[n 0.864 Max 0.881

(.

10/20/97 0.911422 10123197 0.911288 10/26/97 0.910823 10/29/97 0.91067 11/1/97 0.911033 11/4/97 0.910283 11/7/97 0.911725 11/12/97 0.911467 11/15/97 0.909684 11/18/97 0.90965 11/21/97 0.911011 11/22197 0.911955 11/25/97 0.910925 12/1/97 0.911709 S

129 0.911567 12/7/97 0.912488 12/11/97 0.910597 12/13/97 0.9117 12/17/97 0.910822 12/22/97 0.910759 12/25/97 0.910494 12/30/97 0.912247 1/3/98 0.910127 1/4/98 0.910641 115198 0.912216 1/8198 0.909723 1/11/98 0.911636 1/14/98 0.912836 3/10/98 0.91232 3/13/98 0.911975 3/16/98 0.913333 3/19198 0.912547" 3/23/98 0.913395 3/26/98 0.912127 3/29/98 0.912964 4/1/98 0.912778 4/5198 0.91228 4110198 0.911908 4113198 0.912256 4/16/98 0.912156 4119/98 0.911492 4/22198 0.91145 4/25/98 0.910739 4/29/98 0.911322 5/3198 0.911391 5/7/98 0.910847 5/10198 0.91009 5/11/98 0.910342 5/14198 0.910952 1.030964 1.030804 1.031504 1.03072 1.030843 1.031318 1.030979 1.031886 1.030567 1.031071 1.031691 1.030923 1.030846 1.030873 1.030754 1.032515 1.030948 1.03124 1.031321 1,031114 1.031079 1.031204 1.030744 1.031027 1.031251 1.030582 1.031042 1.031876 1.030635 1.030465 1.030545 1.031612 1.030638 1.030668 1.031743 1.030843 1.030837 1.030738 1.031035 1.031139 1.031101 1.03124 1.031131 1.03117 1.031085 1.030593 1.030934 1.030826 1.031115 1.0258 1.026233 1.026489 1.025297 1.025133 1.0246 1.025017 1.027275 1.025617 1.024977 1,026167 1.025257 1.025175 1.024621

1. 024402 1.025866 1.02488 1.024343 1.024885 1.024749 1.025811 1.024402 1.025146 1.024928 1.026291 1.025653 1.024733 1.025988 1.026682 1.02719 1.02697 1.02756 1.027459 1.027582 1.026811 1.027709 1.02802 1.027673 1.026883 1.027283 1.028304 1.027254 1.027384 1.027112 1.027426 1.028301 1.026926 1.027685 1.028837 0.891634 0.901528 0.890882 0.901085 0.888056 0.899439 0.895056 0.902863 0.894803 0.902918 0.895811 0.903047 0.894061 0.902893 0.883359 0.8974.13 0.895483 0.902583 0.89582 0.902735 0.888367 0.899689 0.893318 0.902637 0.894706 0.902815 0.89573 0.90372 0.897075 0.904321 0.884905 0.898697 0.895936 0.903266 0.895503 0.903602 0.894197 0.90251 0.89547 0.903114 0,892239 0.901367 0.894903 0.903575 0.896021 0.903074 0.895336 0.902989 0.888186 0.900201 0.895203 0.902463 0.894914 0.903275 0.886628 0.899732 0.888901 0.90061 0.887959 0.899967 0.887136 0.900235 0.882411 0.897479 0.885189 0.899292 0.885984 0.899055 0.88395 0.898457 0.884239 0.898508 0.88365 0.897965 0.885594 0.898751 0.888837 0.899546 0.885285 0.89872 0.882511 0.897001 0.885683 0.898566 0.885211 0.898475 0.886475 0.898899 0.885647 0.898519 0.884961 0.897904 0.889184 0.899637 0.886653 0.898498 0.881017 0.895985 1.028382 0.877 1.028519 0.876 1.028997 0.874 1.028009 0.878 1.027988 0.878 1.027959 0.878 1.027998 0.878 1.029581 0.872 1.028092 0.878 1.028024 0.878 1.028929 0.874 1.02809 0.878 1.028011 0.878 1.027747 0.879 1.027578F0.880 1.029191

'0.873 1.027914 0.879 1.027792 0.879 1.028103 0.878 1027932 0.879 1.028445 0.876 1.027803

.0.879 1.027945 0.879 1.027978 0.878 1.028771 0.875 1.028118 0.878 1.027888 0.879

.1.028932 0.874 1.028659 0.876 1.028828 0.875 1.02B758 0.875 1.029586 0.872 1.02C049 0.874 1.029125 0.874 1.029277 0.873 1.029276 0.873 1.029429 0.872 1.029206 0.873 1.028959 0.874 1.029211 0.873 1.029703 0.871 1.029247 0.873 1.029258 0.873 1.029141 0.873 1.029256 0.873 1.029447 0.872 1.02893 0.874 1.029256 0.873 1.029976 0.870

5/17198 0.910364 1.030179 1.027799 0.898651 0.899507 1.028989-0.874 5/22/98 0.910347 1.030757 1.026973 0.889342 0,899845 1.028865 0.875 5/26/98 0,91137 1.03147 1.027937 0.852531 0.89695 1.029704 0.871 5/29/98 0.910369 1.030909 1.026817 0.889384 0.899876 1.028863 0.875 611198 0.909453 1.03012 1.026942 0.892609 0.901031 1.028531 0.876 614/98 0.909902 1.030756 1.027982 *0.886519 0.898211 1.029369 0.873 616/98 0.90989 1.029693 1.02796 0.890151 0.90002 1.028827 0.875 619198 0.910039 1.030587 1.028685 0.884417 0.897228 1.029636 0.871 6112/98 0.910494 1.030035 1.027076 0.891387 0.900941 1.028558 0.876 6115198 0.910162 1.03138 1.028188 0.883086 0.896624 1.029784 0.871 6/18/98 0.909025 1.029335 1.027693 0.893209 0.901117 1.025514 0.876 6/24/98 0.909612 1.029654 1.027488 0.892298 0.900955 1.028571 0.876 6/27/98 0.907627 1.028976 1.027696 0.895759 0.901693 1.028336 0.877 6/30/98 0.91223 1.03034 1.027903 0.885675 0.898952 1.029122 0.874 7/3198 0.910986 1.029965 1.027286 0.890401 0.900693 1.028626 0.876 7/7/98 0.909958 1.030337 1.026929 0.891442 0.9007 1.0283633 0.876 7M10/98 0.910928

.1.0305 1.027306 0.888453 0.89969 1.028903 0.874 7114/98 0.909859 1.030626 1.026899 0.890714 0.900287 1.028763 0.875 7/17198 0.911722 1.031477 1.027491 0.883862 0.897792 1.029484 0.872 7/18/98 0.910188 1.030573 1.028039 0.886537 0.898362 1.029306 0.873 7/20M98 0.910138 1.030657 1.Q27247 0.889126 0.899632 1.028952 0.874 7/23/98 0.909694 1.02988 1.027618 0.890872 0.900283 1.028749,

0.875 7130/98 0.910609 1.030209 1.02726 0.890093 0.900351 1.028735 0.875 8/2/98 0.910362 1.030421 1.027226 0.889794 0.900078 1.028823 0.875 814198 0.908998 1.030784 1.028193 0.88642 0.897709 1.029489 0.872 8/7/98 0,908611 1.030201 1.027824 0.890059 0.899335 1.020013 0.874 8/7/98 0.908611 1.030201 1.027824 0.890059 0.899335 1.029013 0.874 8/10/98 0.908501 1.030065 1.027945 0.890342 0.899421 1.029005 0.874 8/13/98 0.910428 1.030512 1.027503 0.888361 0.899394 1.029008 0.874 8116198 0.908901 1.029982 1.027667.0.891162 0.900031 1.028825 0.875 F 8/19)981 0.91012 1.032095 1.028763 0.878812 0.894466 1.03r429LF o0a.

8/25/98 0.910642 1.029834 1.026611 0.893428 0.902035 1.028223 0.877 8127198 0.909814 1.029745 1.027177 0.89287 0.901342 1.028461 0.876 8/30/98 0.911269 1.029857 1.027198 0.890812 0.90104 1.028528 0.876 9/2/98 0.909108 1.029626 1.0272 0.893678 0.901393 1.028413 0.876 9/5/98 0.910936 1.030629 1.028486 0.884064 0.8975 1.029558 0.872 918/98 0.910035 1.030021 1,027304 6.89099 0.900513 1.028663 0.875 9/12198 0.909498 1.030026 1.027056 0.892428 0.900963 1.028541 0.876 9115198 0.910383 1.031121 1.026021 0.891484 0.900933 1.028571 0.876 9118/98 0.910642 1.032304 1.027159 0.882942 0.896792 1.029732 0.871 9/21/98 0.907328 1.030548 1.02774 0.890578 0.898953 1.029144 0.873 9/24/98 0.910292 1.030834 1.026976 0.889101 0.899696 1.028905 0.874 9/27/98 0.909814 1,03012 1.026514 0.893595 0.901705 1.028317 0.877 9130/98 0.909489 1.030057 1.026447 0.894559 0.902024 1.028252 0.877 1013/98 0.909553 1.030168 1.026399 0.894256 0.901904 1,028284 0.877 10/6/98 0.910625 1.030907 1.027518 0.886734 0.898679 1.029213 0.873 10/9/98 0.909117 1.030115 1.027228 0.892079 0.900598 1.028672 0.875 10/12/98 0.909706 1.030192 1.026512 0.89352 0.901613 1.028352 0.877 10/15/98 0.909517 1.030408 1.026452 0.893342 0,90143 1.02B43 0.877

10118198 0.909917 1.030926 1.027442 0.887539 0.898728 1.029184 0.873 10121198 0.910431 1.030917 1.026481 0.890514 0.900472 1.028699 0.875 10/24198 0.909686 1.030992 1.026521 0.8909 0.900293 1.028757 0.875 10127198 0.909844 1.031127 1.026245 0.891103 0.900474 1.028686 0.875 10/30/98 0.909936 1.030813 1.02803 0.885908 0.897922 1.029422 0.872 11/2198 0.91047 1.030654 1.026923 0.889753 0.900111 1.023789 0.875 1115/98 0.909503 1.030226 1.027067 0.891864 0.900684 1.028647 0.876 1118/98 0.908368 1.029756 1.027234 0.89397 0.901169 1.02,3495 0.876 Loop 32 Min 0.868 Max 0.880

Prairie Island 2 Data from remote monitoring program - found under LEFMLOGS Loop A S

L S/L 9/21/98 0.8798 1.0181 1.0369 0.9296 0.9047 1.0275 0.881 10/6198 0.8708 1.0166 1.0463 0.9110 0.8909 1.0315 0.864 Loop B S

L S/L 12/4197 0.9116 1.0308 1.0244 0.8971 0.9043 1.0276 0.880 8/19/98

,0.9101 1.0321 1.0288 0.8788 0.8945 1.0304 0.868

Plant Name:

Beaver Valley Unit 1 Feedwater Measurement System:

LEFM,/

Installation Geometry:

10 Diameters Downstream from a Header Non-planar bend 4 Diameters Upstream Chordal Meter Measurement Error = 0.01%

.OA5 0

M5 ER-262 Rev. 0 Count on Caldon Appendix F

TYPICAL PIPING CONFIGURATION AND LEFM LOCATION BEAVER VALLEY UNIT 1 SKETCH SKRSH-02A.DWG ER-262 Rev. 0 Count on Caldon Appendix F

Beaver Valley 1 09:49:40 2001/10/22 Configuration Files ALARM.INI FAT.INI HYDRAULI.INI METER.INI PARAMETR. INI P CONFIG.INI PROPERTY.INI SETUP.INI Setup Files Setapul.txt Setapu2.txt 2001/10/12 2001/03/22 2001/05/08 2001/07/13 2001/05/08 2001/04/17 2001/03/22 2001/05/08 19:31:40 15:07:12 06:40:22 17:35:22 17:26:30 13:30:02 15:20:40 06:19:42 FFFF909D FFFFF185 FFFF820A FFFFOEFF FFFC946F FFFF6881 FFFFF97A FFFFAA6B 2001/04/17 16:21:00 FFFEOF61 2001/02/21 13:44:14 FFF89974 Beaver Beaver Beaver Beaver Beaver Beaver Beaver Beaver Beaver Beaver Valley Valley Valley Valley Valley Valley Valley Valley Valley Valley 1 Current Flow:

1 Average Flow:

1 Maximum Flow:

1 Minimum Flow:

2 Deviation Flow:

11 2

1 1

Current Temp:

Average Temp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

62.30 62.36 62.52 62.16 0.07 434.0 434.0 434.0 433.9 0.0 NORMAL NORMAL 11.171 11.782 11.814 11.744 0.013 0.12 Beaver Valley 1 Current System Status:

Beaver Valley 1 Minimum System Status:

Beaver Beaver Beaver Beaver Beaver Valley Valley Valley Valley Valley 1 Current Mass Flow:

1 Average Mass Flow:

1 Maximum Mass Flow:

1 Minimum Mass Flow:

I Deviation Mass Flow:

Beaver Valley 1 Uncertainty:

Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1

1 1

1 1

1 1

1 1

1 1

1 1

1 Current Flow:

Average Flow:

Maximum Flow:

Minimum Flow:

Deviation Flow:

Current Temp:

Average Temp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

Current Press:

Average Press:

Maximum Press:

Minimum Press:

Deviation Press:

62.30 62.36 62.52 62.16 0.07 434.0 434.0 434.0 433.9 0.0 1090.60 1091.02 1093.03 1089.55 0.02

Meter 1 Current Meter Status:

NORMAL Meter I Minimum Meter Status:

NORMAL Meter 1 Current Mass Flow:

11.771 Meter 1 Average Mass Flow:

11.782 Meter 1 Maximum Mass Flow:

11.814 Meter I Minimum Mass Flow:

11.744 Meter I Deviation Mass Flow:

0.013 Meter I Uncertainty:

0.12 Path 1 Path 2 Path 3 Path 4 Meter I Current Variance:

122656.34 86921.20 101972.74 77350.00 Meter 1 Average Vnorm:

1.0594 1.0682 0.9673 0-.8175 Meter 1 Current Vnorm:

1.0569 1.0638 0.9704 0.8246 Meter 1 Maximum Vnorm:

1.1004 1.0852 0.9852 0.8428 Meter 1 Minimum Vnorm:

1.0179 1.0547 0.9483 0.7852 Meter I Deviation Vnorm:

0.015 0.005 0.006 0.011 Meter 1 Benchmark Vnorm:

1.0585 1.0679 0.9678 0.8179 Meter 1 Limit % Vnorm:

3.00 3.00 3.00 3.00 Meter I Average Gain:

63.08 62.65 64.56 66.38 Meter 1 Current Gain:

63.07 62.59 64.68 66.56 Meter 1 Maximum Gain:

63.42 63.09 64.73 66.76 Meter 1 Minimum Gain:

62.68 -

62.33 64.38 65.99 Meter 1 Deviation Gain:

0.12 0.12 0.06 0.13 Meter 1 Limit Gain:

75.00 75.00 75.00 75.00 Meter 1 Current Gain Up:

60.99 62.41 63.97 65.07 Meter 1 Current Gain Down:

65.07 62.88 65.23 68.05 Meter 1 Current TPGain Up:

63.97 64.13 64.13 64.13 Meter I Current TPGain Down:

64.13 64.29 63.97 64.29 Meter 1 Average S/N Ratio:

48.94 50.00 34.90 30.33 Meter I Current SIN Ratio:

49.13 50.02 34.75 29.80 Meter 1 Maximum S/N Ratio:

49.85 50.82 35.37 30.85 Meter 1 Minimum S/N Ratio:

48.25 49.06 34.52 29.66 Meter 1 Deviation S/N Ratio:

0.29 0.33 0.14 0.25 Meter 1 Average TDown:

417993 661915 662417 419235 Meter 1 Current TDown:

418011 661957 662429 419240 Meter 1 Maximum TDown:

418066 661993 662487 419294 Meter 1 Minimum TDown:

417930 661823 662339 419177 Meter 1 Deviation TDown:

22 26 26 18 Meter I Current TPTDown:

4000452 4000454 4000452 4000449 Meter 1 Average DeltaT:

2542.9 4760.7 4311.6 1962.6 Meter 1 Current DeltaT:

2534.9 4737.2 4321.8 1978.1 Meter I Maximum DeltaT:

2643.3 4839.5 4384.9 2019.7 Meter 1 Minimum DeltaT:

2444.8 4691.2 4230.4 1883.9 Meter 1 Deviation DeltaT:

36.3 24.3 27.8 27.0 Meter 1 Current TPDeltaT:

2.0

-2.9 1.7 4.4 Meter 1 Current Path Status:

NORMAL NORMAL NORMAL NORMAL Meter I Minimum Path Status:

NORMAL NORMAL NORMAL NORMAL Meter I Average Reject %:

0.9 0.3 0.3 0.6

Meter 1 Current Reject %:

1.2 0.0 0.1 0.4 Meter 1 Maximum Reject %:

2.2 1.2 1.2 1.8 Meter 1 Minimum Reject %:

0.0 0.0 0.0 0.0 Meter I Deviation Reject %:

0.5 0.2 0.2 0.4 Meter 1 Incoming Samples:

719 719 719 719 Meter I Number Failed Rejects:

0 0

0 0

Alarm Log Events

HYDRAULI.ini REM Sound Velocity Ratio to Nominal DEFAULTCFRATIO1:,0.9998,1.0002,1.0004,1.0000 REM Nominal Sound Velocity for the Speed of Sound Tests SOUNDVELOCITYNOM1:,50300 REM Averaging period for the Velocity Profile Benchmark Calculation MAXN:,720 REM Velocity Profiles used to evaluate the profile test DEFAULTVELOCITYI:,1.1080,1.0894,0.9448,0.7765 REM Profile Factor Coefficients PROFILEFACTORCOEFAO:, 1.0039E+000

(,...

Page I

t--

%13 =N17= E0 Plant Name:

Feedwater Measurement System:

Installation Geometry:

Beaver Valley Unit 2 LEFM,/

6 Diameters Downstream from a Header Two Non-planar Feeds Upstream Chordal Meter Measurement Error= 0.01%

(.

-0.5 0

0.5 tcrMtO.R.Ous ER-262 Rev. 0 Count on Caldon AAppendix F

~12 0

~

0 A

SKETCH SKRSH-02B0.DG TYPICAL PIPING CONFIGURATION I

I II 7

I AND LEFMd LOCAMON BEAVER VALLEY UNIT 2 ER-262 Rev. 0 Count on Caldon Appendix F

Beaver Valley Unit 2 09:37:29 2001/10/22 Configuration Files ALARM.INI FAT.INI HYDRAULI.INI METER.INI PARAMETR.INI P CONFIG.INI PROPERTY.INI SETUP.INI Setup Files Setapul.txt Setapu2.txt Setapu3.txt Setapu4.txt SetapuS.txt Setapu6.txt Setapu7.txt Setapu8.txt Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit 2001/06/18 2001/03/23 2001/06/18 2001/07/13 2001/06/18 2001/05/02 2001/03/23 2001/07/05 2001/06/18 2001/05/08 2001/03/23 2001/03/23 2001/06/18 2001/03/23 2001/03/23 2001/03/23 09:52:56 15:40:46 12:13:02 17:35:50 12:15:56 10:02:04 15:55:34 15:21:36 15:00:40 13:56:36 15:25:32 15:25:32 15:01:34 15:25:32 15:25:32 15:25:32 FFFF04 DE FFFFAA26 FFFF49AE FFFC458F FFFC7630 FFFD81CB FFFFD6AC FFFE7200 FFFDFBO4 FFFE1903 FFFE1904 FFFE1904 FFFDFAE5 FFFE1904 FFFE1904 FFFE1904 61.31 61.33 61.43 61.23 0.03 432.9 432.9 432.9 432.9 0.0 NORMAL NORMUL 22 2

2 2

(:. :9.

Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit Beaver Valley Unit 2

2 2

2 2

Current Flow:

Average Flow:

Maximum Flow:

Minimum Flow:

Deviation Flow:

Current Temp:

Average Temp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

Beaver Valley Unit 2 Current System Status:

Beaver Valley Unit 2 Minimum System Status:

Beaver Beaver Beaver Beaver Beaver Valley Unit Valley Unit Valley Unit Valley Unit Valley Unit 2

2 2

2 2

Current Mass Flow:

Average Mass Flow:

Maximum Mass Flow:

Minimum Mass Flow:

Deviation Mass Flow:

11.593 11.597 11.610 11.578 0.006 0.10 Beaver Valley Unit 2 Uncertainty:

Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1 Current Flow:

1 Average Flow:

1 Maximum Flow:

I Minimum Flow:

1 Deviation Flow:

61.31 61.33 61.40 61.23 0.03 432.9 432.9 432.9 432.9 0.0 11 1

1 I

Current Temp:

Average Tamp:

Maximum Temp:

Minimum Temp:

Deviation Temp:

Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter Meter 1 Current Press:

1 Average Press:

1 Maximum Press:

1 Minimum Press:

1 Deviation Press:

1 Current Meter Status:

1 Minimum Meter Status:

1 Current Mass Flow:

1 Average Mass Flow:

I Maximum Mass Flow:

1 Minimum Mass Flow:

1 Deviation Mass Flow:

1 Uncertainty:

Path 5 Path 6 Path 7 Meter 1 Current Variance:

112477.57 135183.91 125798.

Meter 1 Average Vnorm:

1.1506 1.1063 0.9360 Meter I Current Vnorm:

1.1490 1.1014 0.9410 Meter 1 Maximum Vnorm:

1.1839 1.1236 0.9588 Meter 1 Minimum Vnorm:

1.0892 1.0869 0.9177 Meter 1 Deviation Vnorm:

0.014 0.007 0.007 Meter 1 Benchmark Vnorm:

1.1522 1.1068 0.9353 Meter 1 Limit % Vnorm:

0.50 0.50 0.50 Meter 1 Average Gain:

60.13 59.16 69.79 Meter 1 Current Gain:

60.25 59.22 69.73 Meter 1 Maximum Gain:

60.30 59.31 69.96 Meter 1 Minimum Gain:

59.93 58.99 69.67 Meter I Deviation Gain:

0.07 0.06 0.06 Meter 1 Limit Gain:

76.00 76.00 76.00 Meter 1 Current Gain Up:

60.37 59.42 69.78 Meter 1 Current Gain Down:

60.05 58.95 69.62 Meter 1 Current TPGain Up:

64.60 64.60 64.76 Meter 1 Current TPGain Down:

64.76 64.44 64.60 1087.95 1087.57 1088.36 1056.85

.0.01 NORMAL NORMAL 11.593 11.597 11.610 11.578 0.006 0.10 Path 1 Path 2 Path 8 109352.58 119133.39 45 128769.61 0.7408 0.9327 0.7313 0.7395 0.9335 0.7169 0.7828 0.9511 0.7808 0.7124 0.9191 0.6994 0.012 0.006 0.014 0.7404 0.9322 0.7307 0.50 0.50 0.50 69.89 63.03 56.50 69.95 63.10 56.52 70.12 63.20 56.64 69.75 62.89 56.38 0.06 0.05 0.05 76.00 76.00 76.00 69.62 63.03 55.66 70.09 63.03 57.23 64.60 64.76 64.44 64.44 64.44

. 64.92 Path 3 125851.56 1.0980 1.0987 1.1167 1.0813 0.006 1.0983 0.50 61.14 61.11 61.25 60.99 0.06 76.00 61.46 60.68 64.76 64.44 Path 4 102142.39 1.1240 1.1161 1.1549 1.0757 0.014 1.1246 0.50 66.59 66.56 66.80 66.39 0.07 76.00 65.86 67.11 64.60 64.60

Meter 1 Average S/N Ratio:

25.63 55.07 57.63 31.86 75.85 82.27 22.36 92.80 Meter 1 Current S/N Ratio:

25.66 55.15 58.06 31.97 75.77 82.67 22.38 92.72 Meter 1 Maximum S/N Ratio:

25.86 55.53 58.26 32.25 77.14 83.15 22.54 93.74 Meter 1 Minimum S/N Ratio:

25.44 54.69 57.13 31.56 74.80 81.50 22.17 91.74 Meter 1 Deviation S/N Ratio:

0.07 0.15 0.23 0.13 0.35 0.33 0.06 0.31 Meter I Average TDown:

378926 635665 634623 378186 383409 634129 634513 378401 Meter 1 Current TDown:

378930 635663 634624 378194 383410 634138 634499 378394 Meter 1 Maximum TDown:

378970 635710 634670 378237 383479 634184 634563 378450 Meter 1 Minimum TDown:

378874 635618 634573 378147 383366 634081 634459 378343 Meter I Deviation TDown:

17 19 17 16 18 19 19 18 Meter 1 Current TPTDown:

4500402 4500402 4500402 4500402 4500508 450050.7 4500508 4500507 Meter 1 Average DeltaT:

1733.5 4044.1 4761.3 2634.8 2689.9 4790.6 4052.9 1713.3 Meter I Current DeltaT:

1729.9 4046.1 4762.8

.2615.4 2685.5 4767.7 4073.1 1725.8 Meter 1 Maximum DeltaT:

1832.5 4125.9 4840.1 2709.1 2768.7 4867.7 4153.0 1829.6 Meter 1 Minimum DeltaT:

1666.1 3986.5 4690.9 2522.6 2547.1 4707.4

.3975.7 1638.9 Meter 1 Deviation DeltaT:

28.2 27.1 26.6 32.8 32.7 32.2 29.2 33.2 Meter 1 Current TPDeltaT:

0.1 0.7

-0.0

-0.1 2.2

-0.2 0.2 4.5 Meter 1 Current Path Status:

NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL Meter 1 Minimum Path Status:

NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL NORMAL Meter 1 Average Reject %:

0.1 0.1 0.1 0.2 0.1 0.1 0.2 0.1 Meter 1 Current Reject %:

0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 Meter I Maximum Reject %:

1.2 0.7 0.8 1.0 1.2 0.8 0.8 1.1 Meter 1 Minimum Reject %:

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Meter 1 Deviation Reject %:

0.2 0.1 0.2 0.2 0.2 0.1 0.2 0.2 Meter 1 Incoming Samples:

719 719 719 719 719 719 719 719 Meter 1 Number Failed Rejects:

0 0

0 0

0 0

0 0

Meter I Deviation Flowz 0.01 Meter 1 Currint Tempt 434.2 Meter 1 Average Tempt 434.2 Meter I Maximum Temp:

414.3..

Meter 1 Minimum Tempt 414.2' Meter 1 Deviation Temp:

0.;0 meter i Current "ireai 115.01 Meter I Average Pteas 10)4.61 Meter 1 Maxitnum Press:

41075.73 Meter I Minimum Preent 10.73.57 "

Meter 1 Deviation PreSet 0.01 Meter 1 Current Hieter. Statue:

I(OtMAL Meter 1 Minirtum Meter Status; NOIUMAL.

Meter 1 Current Mass Plow:

11.760-Meter 1 Average Mass Plow:

11.779 Miter I Maximum Maca Flow:

11794-Meter 2 Minimum tiesa Flow:

iiv 0do.

Meter I Deviation Masa Flow:

0;005' Meter I Uncertainty:

0.10 Path 1 PAth 2 Path 3 Path 4 Path 5 Path 6 Path 7 Path B Meter I Current Variance:

101724.19 128352,17 119290.99 120135.55 127986.56 IS9406.06 142563.78 137988.53 Meter 1 Average Vnorm:

0.7338 0.9276 1.1020 1.1252 1.1518 1.1157 0.9326 0.7108 Meter I Current Vnorm:

0,7440 0.9336 1.0970 1.1115 1.1473 1.1037 0.9414 0.7349 Meter I Maximum Vnormt"!

0.7731 0.9514 1.1208 1.197n 1.1909 1.1363 0.9505 0."563 Meter 1 Minimum Vnorm:

0.6977 0.9080 1.0792 1:0767 1.1162 1.0987 0.9136 0.6771 Meter, I Deviation Vnorm:

0.012

.0.007 0.007 0.015 0.014 0.007 0.007 0.05 Meter I Benchmark Vnorm:

0.7348 0.9282 1.1014 1.1244

-'1.1507 1.1154 0.9330 0.7200 Meter 1 Limit % Vnorm:

0.0" 0.50 0.50 0.50 0.50 0.50 0.50 0.50 Meter I Average Gain:

66.908 64.51 62.39 64.05 62.38 60.2S 67.51 57.07 Meter 1 Current Gain:

67.02 64.41 62.47 64.03 62.49 60.24 67.52 57.01 Meter 2 Maxiinum Gaini 67.14 64.68 62.52 64.22 62.63 60.1 9 61.70 57.22 Meter I Minimum Oain:

66.81 64.38 62.24 63.83 62.14 60.01 67.31 56.93 Meter 1 Devibtion Gain:

0.0 0.05 0.05 0.06 0.00 0.07 0.06 o005 Meter I Limit Gains 76.00 76.00 76.00 76.00 76.00 76,00 76.00 76.00 Meter 1 Current Gain Up:

66;48 63.97 62.72 62.88 63.50 60.21 67.90 56.13 Meter I Current Gain Down:-

67 42 64.60 62.09 65.07 61.31 60.05 66.95 S7.86 Meter 2 Current TP~aih Ups 64;60 64.66

'64.76 64.76

.64.44 64.44 64.60 64.44 Meter I Current TPOairn Down:

64.60 64.60 64.60 64.60 64.60 64.44 64.60 64.76 Meter I Average S/N Ratio:

37.21 48:21 50.49 41.69 54.93 74.50 28.73 85.88 Meter 1 Current i/N Ratio:

37.33 48.04 50.44 41.75 54.91 75.06 28.81 04.86

Beaver Valley Data taken from commissioning and from plant personnel during the velocity profile alarm Unit 1 5/14101 10/22/01

-0.861136 1.1080 1.0594

-0.339981 1.0894 1.0682 0.33998 0.9448 0.9673

.0.86114 0.7765 0.8175 S/L 0.926 0.922 Unit 2 6/18/01 10/22/01

-0.861136 0.7263 0.7361

-0.339981 0.9301 0.9344 0.33998 1.1089 1.1022 0.86114 1.1385 1.1373 S/L 0.915 0.920 Path 1 Path.2 Path 3 Path 4 Path 5 Path 6 Path 7 Path 8 6/18/01 10/22/01 0.7338 0.7408 0.9276 0.9327 1.1020 1.0980 1.1252 1.1240 1.1518 1.1506 1.1157 1.1063 0.9326 0.9360 0.7188 0.7313

'I

..b.

,q, (4

'1 III A..

N

~

• 5~vP7*

t.

IL W.3*.I I 1031104 20403

£R FU3WATOW3RIN. Ali A 4 W.

MAL aPA11MG CCON1TK~ksr-M-MA$4c ag t43-D CK 100. L IN'. 3'NL 4

ALL 6mw..A.p~

.0

.Q pm

• o

t-7.5

[-

ItJ~

1 f*1

/

La

~3 A,

0 fl't 4404-I 9!a0'4.PIO~

IAldP~1TI WArid5VA?4 S -

-90%'S-W404' Ilb. St.DL OM

~0~4qT iffl 5 A IlL 50-0

't RL ISW.r - W" P0fTI

• OIL aq STL.ML# F1uL OLaw

,Is-3.-.a W, lu.uoam.owT~s P"

02"k4 UG 40.IG' UobfoIIV0 S,

I'mqIWU m i*

a d L.i TO

SS.t.GU COMOIrIOMS

-"0*(T~

F jM J#

ADCP-TO 9U51 4 rl Wi A

AS EuzM6 uj bal-rW WAIt MflDt I,

$414134L ASTMA-fI.. & kCtPrCLJ sle 40 kALL -r!5Nam.

j7'7 KATtRIAL ASrM AIOG 5?6' 5OWZ

V I,5L. TH'K iO..714' tAATGA1AL AST.

AJO 4

ISS

• q'P-2-0204, 15~

PW re' CW IOUI f

.skI.l

.l &I Il. W-0.

1's-F-b2ut,

'/ PIS Fb MIS CHAS!I ISA 14A PL A]. E'.W 0

'10 otl%40 MS.4 Pb, P

,5 C

UK !'FLAM AT ft..o5`0' k'I.5-2F-lJL159 14 F4iJi

'Cbbh.WW 10')! ISA PA.

1,!1 k

3 k l.O*

"0144*P201t~40 1154* P*1J P11 b.Ia IOUII IhA P;LASi.AT A. 30:O 41"I.P-%oIl,-

bbS 4 F PPI,14.

ao-.

405 PLA A

PT W-

-7 qlS-

-;- 20101 WE1PtJAT 9"I"*

BeTM PI

["Ath 1:.I4 20o234 FW OY5 STUM*

I.

46l-

• r*A"Wo'J 4 W*X=TlO4 IIOLW',

q1G5-F-llZSq? FP. SISTRA"UM 4607-FAWe CA'noJ 4 !EICTIO'4 *SOIr5Ely q7sr-F,,23qB-Fw sriM* -um 4508.-

eAI4ICAT)OPJ lEWCTI-N I-I0ET 1C

.q7'-$-?ot*qq 4i U"(mM.-uI 4601-zMPIIeAO Sllro IJ.. omlITIc FOR8EIN PRINTS "l1&b-:LoA5 COfE VtLCA.

iWt.i CO*"

O'L.

VALMe VVV-610,1l2Ma.

o 4 54C) 154.s - 70 WZ4 I

baM

,,T~83.E'~k,5

JcGi - -00*

L.C* & 1.0, 4210, 4 a.440.

17 1 1 9511 IWCHJ.L VA.LI.Fw.4VIFW-VsLFW.V4714FAV-Vf.

I.U. 0l5 SI o1 WI. WFS"III Ir 10140W 011 L ROh o4115 1

SWI M

aW i

5 O5115 LOO 146401 Sit 7,TR*h.*4I-

."s, 2

I

I.

1.0 L

t d+w.

A--l

-i

.i 0

0%I~

it t*4

.4-.

.Jo LME4004.6 1580O F4WwC r'tm

.5r4ý1

.N,>K-40 MS t7 "~S 4V

OIL, N~~~>*Am n*c sn,4M.00 n 4.A

-gotl U.6U-0-M'l Z~m.

=

L ill.~a~

rb~mhi, P"~0.%- EAwm To 04*imool4 00405 XA FEUEVWATN 0.1.

14-A fdW pIoS..indO P104 1w*. 140 Fr MAX. OPEROWCUITlS~fWT&A4 AL 01.i4 W4L U 01 L

M 09

3.

'S v,

• " 2.

I

r F!

F E

Cam=>

in, U~4

-t ME maIa-iFEWTRSSE 12.., IQ

9.

a' in.5af372.&

-A--

I It to qS 0

1 4

3 9-9OZA.

-oldj

-~-~

L.-

'I L

H 6

-F 1

0 IP:A4 Im kip Jý.

(P N

I cvw i F IL Ila I; I it'd!

H ~

6 wU\\

ii, em

• .km 4=4 =7 K

"M OM-PORTION -OF --

RWN R

FNonZ SAET iE G

E 0

c S

A USl -

Ia

- I-kwl-.

North Atlantic &-w s.*. cPn=,

DTETAILS PI0-I-F W-0206BG

• I?

'I IS q

6 7

5 2

ml.m,

.12 I1 IN 4

8 7

6 2

H

£89~O-I It 10 4

I 6

5 4

It 10 9

7 B

5 3

m*-p fl.nnn. fan rm -

a I

1 3

nt-V MAINV

• j'..

o LFT=

LEGEND G isii it musk n

l t

0~fnn n

n

-N G nsmi

-li tf G

I (j) an.n.nc inn -

in -

baift fttbbbft.dn.

ft..

fti nun nu inn.

E 0

k ;jý I B

A II

~

a-ft n-i.,.

I it.

II;

-I~I)

A

/

I iO

-~

S n

itist a-neum--j

-E L i 1

kV_

C l

WDn Ulu1Z 7 A

FPL ENERG Y

&-ftn FEEDWATER SYSTEM DETAILS Q 9 PnD-I-FW-D20GB7 PlO-l-W-208 12 11 Is 9l 6

5 3

I

~--

~

-- ~.

- -~

It I@

q a

7 6

3 2

7o.O

. *.-! o*Tj H

I I

I I

a..e-.in sac K

r---*:

• E

.- *I eli

~

~U~LiES~.

.4

-I-er LLU-HU ISSlmlt nan Sn..

Se Sm~

-,v nml I

¶

,S -

mir

.ae...

we

-m..n. 1 m

.,I-0

• maI m I

Ima IS.

0-"'-'

a-I2-I F

S5-A DO Oa

.w"-

PORTIONS OF TIMS ORAWMJ ARE F.NMEHuCLYE-AR SAFETT IELATED c

A hI.a~e T.

.FPL ENERGY Sabm Mo EMERGE:Y ffTEl SYSTEM JPI12-I-FW-D20688 o,.2 II

]I 9

7 9

5 4

3

• 2

I -

11 to 7

6 5

4 3

-I-I,

7 I-

'I IS 7

6 5

4 3

r's ~ThJJ'!~'

L*T!EEO m~*,w

,'I Sb.I.,.,

i*I~!iS!!

• N, t~tl tW.4StWS~f SiROll 1,1 I.

C*'~LEGEND OK~MI 0!

• ~

~

MEW F

a c

C B

8 OWN 4-T-.I-~-4-=---I ~T4 ~-I.Y~--l A 1,-

A!1

~

SJ~

N**thII A tlanti c A! Ill c

k..

_____North Atlantic *,W sutsc c=*=

MISELLANER SYSVNTEM RAN DICLAENlLTS

&I RIN PlO-I-FW-0206eq It II IS 9

B 5

.4 2

jkz"4,,

12 I I to 9

a 5

.4 2

• U lB 0

6 2

It

.9.G-ZU-z.- -! -O d j a

i~t *.,*.1*

lt*,*b,.

I "r

T17-6 r

p~ ~

~

~

~~-T-

ý -,m,,*

=*

...]

iHE-

"f,..

'-- 4

-e.,

0 me

-. 0 n-qa LEGEND 0==8

-em a

F F

E a

j C

. Fxi r

swimA

.1 North Atlantic a-1 A

67,0 rEIEODTER SYSTEM WET LAY-UP pin-I-FW-020690O t

.111________

5, 4

me -

IM Ml W'a A

North Atlantic

~

sa~a c~m 12 II IsO I

FEEDVATER SYSTEM WET LAY-UI' PlO-2-FW-O2~69~

12 II I,

7

12 II In q

a 7

6 5

4 3

2 I

12 11 In q

a 7

5 H

a i-rwi-9co-26 0

F b

B 9

_ANorthNrhAtlantic r-am. c.oi,.

A i

i f i--f LUO~JF PlO-I-FW-020691

.& =J==.F-. A A.1.

12 II is B

7 S

4 3

12 tl le a

7 6

4 3

Forth American Nuclear Society Intmrnational-Topicar MNeting on NuclelPIantInstrumemation. Control-and Hunman-Machine nfeIcfc Technologies (NPIC&HfIT 2004),.Coliwnbus,.OlDio;. September, 2004 Traceability of Thermal Power Measurements Don Augenstein Herb Estrada Caldon, Inc.

• 1070 Banksville Avenue Pittsburgh, PA 15216

• daugenstein(acaldon.net herbestrada(&@comcast.net Ernie Hauser Caldon, Inc.

1070 Banksville Avenue Pittsburgh, PA 15216 ehausertealdon.net Keywords: traceability, chordal, transit-time, ultrasonic, measurement Part 1 Chordal Ultrasonic Flow Measurements ABSTRACT The traceability of a measurement of nuclear feedwater mass flow by a chordal ultrasonic meter is described, with particular emphasis on the methodology whereby transit time measurements and calibration factors used in the field can be traced to an appropriate standard. This paper is a companion to a paper considering the challenges of traceability of feedwater flow measurements by flow nozzles and venturi tubes.

1. INTRODUCTION A continuous, accurate determination of thermal power is an essential requirement in the operation of a nuclear power plant. Errors in the power determination can cause lost revenue or reduced safety margin-both serious consequences. It is therefore appropriate that the rigor of traceability be applied to each component of the thermal power determination. The desirability to apply rigorous traceability requirements to thermal power determinations is underlined by recent problems with flow instrumentation in nuclear applications.

Traceability is defined as a process whereby a measurement can be related to a standard via a chain of comparisons (International Standards Organization, Reference (1)). Certain requirements apply:

" The standard must be acceptable to all parties with an interest in the measurement and is usually a standard maintained by a national laboratory such as the National Institute of Standards and Technology.

" The chain of comparisons must be unbroken--the field measurement must be connected, by one or more links directly to the standard.

" Every link in the chain involves a comparison that necessarily carries with it an uncertainty. Hence the total uncertainty of the measurement must reflect the aggregate uncertainties of each link of the comparison chain.

" There can be no unverified assumptions in the chain of comparisons; it is clearly not possible rationally to assign an uncertainty to an assumption with no quantitative basis.

In virtually all light water nuclear power plants, thermal power is determined by a power balance around the steam supply. The process involves measuring or otherwise determining the following principal process variables:

(1) The total mass flow into the steam supply, WFw, the total feedwater flow, and the blowdown flow removed from the steam supply (if any), WBD. WBD is returned, purified, via the feedwater system. The third mass flow component of a steam supply mass balance, the steam flow, necessarily equals the difference between the feedwater flow and the blowdown flow in the steady state.+

(2) The specific enthalpy of the water fed to the steam supply, hFw (3) The specific enthalpies of the steam, hs, and the blowdown, hf, exiting the steam supply, (it is generally assumed that the blowdown flow exits the steam supply as saturated liquid)

The blowdown energy flow is typically in the order of V2% of the total power.

This term does not appear in the power balance for BWRs where the blowdown function is carried out by the reactor water cleanup system, or on PWRs that employ once-through steam generators.

Since the objective of the steam supply power balance is to determine the thermal power generated by the reactor core, there are other gains and losses, such as the power added by reactor coolant pumps, that must be accounted. Although the net of these terms rarely aggregates to more than a fraction of 1% of the reactor power rating, diligence requires that they be measured or otherwise determined and that the uncertainties in these measurements be accounted. This paper, however, will focus on the steam supply power balance and more specifically on the traceability of the measurement that affects the thermal power determination most substantially: the mass rate of feedwater flow, WFw.

No instrument measures this variable directly. Two diverse types of instruments are analyzed in this paper and its companion paper:

+ In Boiling Water Reactors, a fourth component, the Control Rod Drive Mechanism flow, delivered to the steam supply, is also accounted. This flow is a very small fraition of the feedwater flow and need not be determined with great precision for a thermal power determination.

(2)

(1) A chordal ultrasonic flowmeter, an instrument that measures the transit times of ultrasonic pulses traveling along chordal paths in a flow element and from these measurements and a measurement of fluid pressure calculates the mass rate of feedwater flow and the feedwater temperature.

(2) A flow nozzle, an instrument that measures difference in static pressures between a tap upstream of the nozzle and a tap in the throat of the nozzle and from this differential pressure measurement and a determination of feedwater density, determines the mass rate of feedwater flow. The density determination is made using a final feedwater temperature measurement, usually from a resistance temperature detector or RTD.

This paper analyzes the traceability chains for the first of these instruments: from its basic measurements--transit times and fluid pressure--to the process variable WFW. It covers explicitly the calibration uncertainties of the flow element(s), including the application of the flow element calibration data taken in a hydraulics facility operating at 100 F and 50 psig to the 430 to 450 F, 1000 to 1200 psig conditions in a nuclear feedwater system at full power. From the analyses of this paper, the reader will obtain an understanding of the factors affecting the traceability and accuracy of a chordal ultrasonic feedwater flow instrument.

2. DISCUSSION The algorithms, and traceability chains for a chordal ultrasonic flow measuring instrument are outlined below. For the principles underlying this type of measurement, the reader is referred to the technical literature (Estrada, Reference (2)).

The discussion is based on an ultrasonic meter having eight paths arranged in two planes of four chords each, at right angles to each other and at a nominal 45' with respect to the axis of the flow element. Because orthogonal paths are paired in four planes parallel to the major axis of the flow element, transverse velocities projected onto each path pair cancel when the velocity measurements of a pair of paths are averaged. Hence the path arrangement makes this eight path flow.meter insensitive to variations in transverse velocity. The chordal arrangement of the paired paths provides axial velocity measurements for each of the four chordal locations. As will be seen, these data can be used to characterize the axial velocity profile.

As derived in Reference (2), the mass flow rate, as determined by a chordal ultrasonic flow meter manufactured by Caldon, Inc. is calculated by (1) the numerical integration of the axial fluid velocity over the pipe cross section to determine the volumetric flow rate, and (2) by multiplying the result by the spatial average of the fluid density. The axial fluid velocity at each of the four chordal locations is determined from the transit times of ultrasonic pulses traveling with and against the direction of flow along the path. Specifically, the mass flow algorithm is:

W- = p PF'F. M ('0D/2) wZ X(At,)

Wh -e p =

. tan(mo)(t, +At, /2h-rdl (1)

Where Wf the mass flow rate through the chordal ultrasonic meter, (lbs/sec)

(3)

p

=

the mean feedwater density, (lbs/cu. in.)

PF

=

the profile (or meter) factor, dimensionless Fa (T)=

the thermal expansion factor. This factor accounts for the difference in internal diameter and -transducer face-to-face distance (Lffi) at operating temperature T versus the temperature at which dimensions were measured To. F (CT) = 1 + 3 a (T-To), where a is the coefficient of thermal expansion of the flow element material in (indin./°F)

ID

=

the internal diameter of the spool piece, (in.)

wi

=

the Gaussian quadrature integration weighting factor for path i, (dimensionless) cpi

=

the angle between path i and a normal to the spool piece axis (deg)

Lfr

=

the face-to-face distance between transducer housings of path i, (in.)

ti

=

the total time of flight of pulse along path i in the direction of flow,(sec.)

tpi

=

the total time of flight along path i against the direction of flow, (sec.)

At1

=

the difference in the total transit times of pulses traveling against the flow and with the flow along path i, (sec.); Ati = t1 - tupi, (sec.)

-i the total of the non-fluid delays of pulses traveling along path i, (sec.) -

T

=

the mean fluid temperature, (°F)

Note that the numerical integration above is carried out for four area segments, although the number of chordal paths is eight. This is because the average of the two velocities measured at each chordal location is, in effect, used to establish the axial fluid velocity at that location, which is the variable to be integrated over the pipe cross section.

To determine the thermal expansion, the fluid temperature is needed. To determine the density, the fluid temperature and its pressure are needed. For a measurement of feedwater flow with a Caldon chordal system, the fluid pressure is measured by a conventional pressure transmitter. The temperature is determined from a measurement of the sound velocity, averaged over the pipe cross section and the fluid pressure. The square of the velocity c of pressure wave propagation through a fluid (the sound velocity) is related to the other state variables for the fluid by the partial derivative of fluid pressure p with respect to density p along a line of constant entropy, s.

C, = ap/apl, (2)

The precision of property tables for steam and water (For example, Reference (3)) is, however, insufficient for an accurate determination of fluid temperature from its sound velocity. Caldon measurement systems therefore rely on a proprietary algorithm, derived from experimental data and confirmed by a large number of comparisons with RTD data (Estrada, Reference (4)).

Expressing the methods employed for determining density and temperature algebraically:

(4)

P=fp (T,p)(3 (3)

T~T (c= p)

(4) 4

c.

=F.1 (M)

[w, Lfs]/[tj + (Ati/2) -ri]

(5) i-I Here FaT) = I + a (T - To)

The function fp for the determination of density is extracted from the ASME steam tables (previously referenced). The function fT is Caldon's proprietary algorithm.

Note that for each set of time and pressure measurements, the procedure for determining temperature and sound velocity is iterative. This is necessary because the determination of sound velocity is itself sensitive to temperature as evidenced by the F.1 (T) term in the equation for the mean sound velocity, cme.

This term accounts for the thermal expansion of the path lengths Lffi from the temperature at which they are measured to the temperature at which the sound velocity is measured.

Fundamentally, the traceability of the mass flow algorithm for a chordal ultrasonic meter requires that a chain of comparisons be constructed for the following elements of that algorithm:

1. The Profile Factor, PF This term essentially characterizes the response of the meter to the axial velocity profile it will see in the field (the numerical integration performed by the meter does not integrate the profile perfectly). For Caldon ultrasonic meters, PF is measured in a hydraulic model of the field application at a certified and traceable hydraulic test facility. Because the flow element to be installed in the field is calibrated, measurement errors in the internal diameter, ID, the path angles, p~j, and, to the extent that they affect the volumetric flow measurement, the path lengths, Lifi are embedded in the Profile Factor and do not affect the accuracy of the field measurement.

2. The time measurements. t, and ti Clearly, the flow measurement accuracy is affected by the accuracy with which the pulse transit times are measured. Furthermore, errors in the measurement of time may small enough not to affect the accuracy of the t measurements, but if they are not reciprocal, can affect the accuracy of the At measurement, which can be seen in the algorithm to be critical to overall measurement accuracy

3. The total non-fluid delays in each path. T The non fluid delays consist of the energy transit delays from the transmitter through the transducer cables, the transducers themselves, the acoustic "windows" which serve as interfaces between the transducers and the flowing fluid, the receiving electronics, to the pulse detection logic. Values for the non fluid delays are, like the internal diameter and path angles, embedded in the calibration factor. However, the non fluid delays in the field may differ from thosein the lab due to different conditions (e.g., different temperatures of the acoustic window) or different components (e.g., cables of longer length); hence (5)

traceability of the field values of the non fluid delays is required. Although mechanisms whereby non fluid delays might change in service are few, some assurance that any change over time is within the uncertainty bounds of the comparison chain is required for measurements in the field.

4. The fluid pressure, p A measurement of fluid pressure is necessary to the determination of temperature and density. [The dimensions of the flow element also change with pressure, but the design is such that the effect on the massflow measurement is negligible.]

The chain of comparisons for the verification of the pressure indications is fairly commonplace, involving allowances for uncertainties in the secondary standards used for calibration, allowances for drift in the transmitter due to environmental and other effects while in service, hysteresis and other non-linear properties of the transmitter, and uncertainties due to the length, configuration and density of the fluid in the transmitter impulse line.

5. The individual path lengths Lfir As noted above, errors in path length as regards the volumetric flow determination are imbedded in the calibration factor (PF). However, knowledge of the absolute value of the path lengths is needed for the iterative computation of sound velocity and fluid temperature (see equations (4) and (5) above). The comparison chain for path length is straightforward, involving, primarily, the secondary standard used for its measurement and observational uncertainties.

However, assurance must be provided that the path length does not change in service, due for example to the deposition of corrosion products.

It is also necessary to verify, by a chain of comparisons, the correlation functions fp, fT, and the thermal expansion terms Fa, and F6 (the latter terms involve the coefficient of thermal expansion for the flow element material). However, these verifications can be performed on a one-time basis. Once accomplished, it is not necessary to reverify the functions to confirm their correctness in a specific measurement.

The weighting factors, wi, and the transverse path locations, which do not appear explicitly in the algorithm but are implicit in the weighting factors, are standard factors for four path Gaussian Quadrature integration (Legendre spacing). Any departure of the spacing from nominal is embedded in the meter calibration factor, that is, the profile factor PF.

Of the five measurements whose traceability is required for the verification of the accuracy of a feedwater flow measurement, the most critical are the Profile Factor, and the time measurements. The balance of this paper will therefore focus on the chain of comparisons used to ensure that the uncertainties in these components of the flow computation remain within the design allowances for their respective chains of comparisons.++

Figure 1 is a diagram of the chain of comparisons used to verify the measurements of ultrasonic pulse transit times. The need to establish traceability for the

+ The traceability chains for the remaining three variables-non fluid delay, pressure, and path lengths-will be made available on the Caldon Website.

(6)

transit time measurement clock (the first link of the chain) is obvious. Additional assurance that environmental or other mechanisms do not degrade this secondary standard in the field is provided by an additional automatic check of the clock against and independent and. diverse time standard (the second link of the chain).

The chain of comparisons for the verification of pulse transit time measurements requires more than a check of the clocks, however. The pulse timing starts with the initiation of transmission, a precisely defined event. But the detection of a pulse after its tran sit and a precise, repeatable measurement of its time of arrival present several challenges. The remaining checks of the comparison chain of Figure I are for the purpose of ensuring that the pulse detection and arrival time measurement comply with the assumptions of the meter's uncertainty analysis.

A detailed description of the means for pulse detection in Caldon's ultrasonic meters is beyond the scope of this paper. Briefly, however, following pulse detection, the zero crossing of a half cycle near the leading edge of the received pulse is. used to define the end point of the transit time (from which process the Caldon Trade Name Leading Edgye Flow Meter derives). The zero crossing is used rather than an amplitude threshold

• because it is insensitive to fluctuations in pulse amplitude due to turbulent refraction and other effects. The accuracy of this measurement can be affected by several factors, the most important of which are the magnitude of noise that may be imbedded in the signal and, particularly with respect to the measurement of the time difference At, non-reciprocal delays in the upstream and downstream pulse transits (the latter effects are brought about by non linearities in the transmission process and by differences in the delays in receiving electronics). As can be seen from the figure, checks are performed to confirm that actual meter performance complies with the assumptions of the analysis that establishes its uncertainty. Specifically, the magnitude of the noise relative to the signal is measured and the reciprocity of signals received by upstream and downstream transducers is confirmed. Several other checks are performed to ensure that statistical assumptions of the uncertainty analysis remain valid.

(7)

Measurement Traceability Standard Factory cTceptance (Automat Field Test Fig. 1: Comparison Chain (8)

The appropriate procedures for establishing the calibrations--meter factors--of instruments used to measure feedwater flow in nuclear and fossil power plants has long been the subject of debate among measurement specialists. The problem is that fluid conditions in the laboratory do not and cannot duplicate fluid conditions in the field-the maximum Reynolds Number achievable in a certified facility is about 3 or 4 million; the Reynolds Number at full power in the feedwater system of a typical nuclear or fossil steam plant is in the order of 10 to 30 million. A chordal ultrasonic meter performs a numerical integration of the axial velocity profile; the axial velocity profile is, in some circumstances, sensitive to Reynolds Number. From the perspective of traceability, one question that must be addressed is how does one verify whatever assumption one has made relative to the behavior of axial profile with increasing Reynolds Number?

Axial profiles are not a function of Reynolds Number only, however. In many applications including some feedwater systems, the thickness of the boundary layer is

.dominated not by fluid viscosity, but by the roughness of the pipe wall. In these applications the profile is insensitive to Reynolds Number, but is affected by the relative roughness. Furthermore, in nuclear and fossil feedwater applications, the flow profile is rarely, if ever, "fully developed"; its shape instead reflects the inertial forces exerted on the fluid by upstream hydraulic features. The specific nature of these features is a third (and in many cases dominant) determinate of profile.

The shape of reasonably symmetrical axial velocity profiles of fluid flowing in the turbulent regime can be numerically described using the inverse power law (Schlichting, Reference (5)). This mathematical representation also allows a profile to be related to the chordal velocity measurements of an ultrasonic transit time meter. (Estrada, Reference (6)). Specifically, the chordal arrangement of Caldon's eight path ultrasonic flow meter, permits the shape of the axial velocity profile to be characterized using the ratio of the average of the velocities measured along the outside (short) chords to the average of the velocities measured along the inside chords. This ratio, called the flatness, can be used to predict the response of ultrasonic meters in both eight path and four path configurations to changes in velocity profile. The flatness ratio defines how flat a flow profile is as compared to other measured profiles. The flatter the velocity profile, the higher the flatness ratio. A perfectly flat profile has a flatness of 1.0. Developedturbulent flow profiles in straight pipe with high relative roughness or low (-10,000) Reynolds number will have a flatness in the 0.75 to 0.8 range. Developed profiles at high (10 million) Reynolds number in pipe of nominal roughness will produce a flatness of about

.0.86; if the pipe is hydraulically smooth a flatness of up to 0.9 is obtained. Downstream of non-planar bends and similar features, flatness can approach 0.98 or more. For nuclear feedwater flow measurements, the flatness for actual profiles measured in service have ranged from 0.81 to 1.01 (in the latter case, the profile was "dished"). About half of them have a flatness of greater than 0.9, which as noted above is the flatness of a fully developed profile in hydraulically smooth pipe. In these cases clearly (and in many of the others), inertial features such as bends and header exits have had a greater influence on profile than either the Reynolds Number or the relative roughness. It should also be noted that many of the inertially dominated profiles are present in locations where conventional wisdom would assert that profiles are fully developed (that is, 10 to 15 diameters downstream of the nearest bend).

(9)

By calibrating chordal ultrasonic meters in a variety of hydraulic configurations of varying flatness, a calibration factor for a high Reynolds Number field application can be determined with very little calibration uncertainty. The calibration process is illustrated in Figure 2 by data for an eight path flow meter for a large nuclear unit. This meter was calibrated in a model of the hydraulic configuration of the unit's feedwater system. Profile flatness -was used to characterize the profile "seen" in this model. The feedwater model configuration was then varied parametrically (e.g., by changing the velocity profile upstream of the most distant hydraulic feature of the model) to provide reasonable assurance that the actual plant flow profiles would be bounded by the calibration data.

The Profile Factor for the field application was selected from the flatness measured in the field and a linear fit of the Profile Factor versus Flatness data collected in the lab. The uncertainty in the fit of the data (in this case +/- 0.04%, 2 standard deviations) is carried as an uncertainty in the meter calibration (Profile Factor). This uncertainty is of course in addition to the other uncertainties of the calibration process (for example, the uncertainties of the hydraulic standard used to perform the calibration).

1.012 1.010 11)08

~~FuI

'~D

[

LrDahwn ioo S1.002

• W

• 1.000 U

0.98.8 0.996 0.994 0.992 0.12 0.83 0.84 0.85 0.88 0.87 0.88 0.89 0.90 Flatný Fig. 2: Dependence of Calibration (Profile Factor) of an Eight Chord Ultrasonic Meter on Profile Flatness; Feedwater Measurement in a Large Nuclear Power Plant (10)

The chain of comparisons required for the traceability of the chordal meter

• calibration-its Profile Factor-is shown in Figure 3. The chain reflects the calibration process described in the preceding paragraphs. It also accounts for the uncertainties of the calibration laboratory itself and for the uncertainties of the time measurements of the electronics used for the calibration test. [Effectively, time measurement uncertainties must be accounted twice, once for their effect on establishing the Profile Factor and once for each measurement made in the field.]

Figure 3 also accounts for another uncertainty. Using data from chordal ultrasonic instruments, it has been observed that axial velocity profiles in nuclear feedwater systems vary in time, sometimes significantly (Reference (6), previously cited). Such variations could potentially alter the flatness enough to call into question the validity of the Profile Factor and its assigned uncertainty. To ensure that this does rnot occur, Caldon chordal meters are equ 'ipped with a "velocity profile deviation" alarm. The alarm alerts the plant operator if the change in flatness has the potential to produce a bias in the profile factor exceeding the allowance for such changes.

(11)

PF= PFo + MP F

dF F = flatness - f (RN, Roughness, upstream hydraulic configuration)

F - (Xi (outside chord velocities))/ (Ji (inside chord velocities))

Flow element calibration at certified hydraulic facility, baseline configuration PF0 (F.) = Qsm -(weight, time, p(T,, P*"))

QUIRM

( tiUM I t.#Uru Facility measurements, weight, time, T, and P, traceable per calibrations traceable to NIST UFM time measurements, ti, tw, traceable to NIST per Figure 1. Length measurements required for flatness calculation) also traceable.

UFM time and length measurements, required for field measurement of flatness, traceable.

Confirmation that change in flatness remains within threshold confirms that shape of axial profile remains within allowance for profile change in meter uncertainty analysis.

4-Tests in varying hydraulic configurations at certified hydraulic facility to establish sensitivity of calibration. PF to flatness F dPF = linear fit (PF, -PF) dF (F, - Fo) 4-Determine flatness Ft in the field; establish PF for field installation Fr o (ZXo-d (outside. chord velocities))/ (Zfr 1 w (inside chord velocities))

PFf = PFo + dPF Ff 0 Set threshold for change in flatness, AFT AFT1= I/dPF x OPF (F) dF Where aPF (F) is the allowance for profile shape uncertainty is the meter uncertainty analysis For subsequent flow measurements in the field, confirm flatness is within threshold:

Fe= (Zi (outside chord velocities))/ (1i (inside chord velocities))

(Fr-Ffo)J < AFT? If so, measurement is valid.

If not, measurement is rejected.

Fig. 3: Traceability of Calibration; Assurance of Applicability of Calibration Data to Field Installation (12)

3. CONCLUSIONS:

Means for establishing the traceability of key variables in the measurement of nuclear feedwater flow using a chordal ultrasonic meter have been described. These means include a quantitative basis for establishing the uncertainty in meter calibration due differences between the calibration established in a certified hydraulic facility at low temperature and the calibration in an operating nuclear feedwater system. Such differences are due not only to the difference in Reynolds Number but to differences in pipe wall roughness and, most importantly, to differences between the hydraulics of the plant and the calibration facility.

4.

REFERENCES:

(1) International Vocabulary of Basic and General Terms in Metrology (VIM),

International Organization for Standardization (ISO).

(2) H. Estrada, "Theory of Ultrasonic Flow Measurement, Gases and Liquids", Caldon Technical Paper TP-44, available on the Caldon Website. The principles of ultrasonic flow measurement and the derivation of the algorithm for a chordal instrument are described in this paper.

(3) For instance, the 1967 ASME Steam Tables.

(4) H. Estrada, "An Assessment of the Integrity and Accuracy of Feedwater Flow and Temperature Instruments" EPRI Plant Performance Improvement Seminar, September, 1996.

(5) Hermann Schlichting, Boundary Layer Theory, Chapter XX, McGraw Hill (6) H. Estrada, "Effects of Velocity Profiles Measured In-Plant on Feedwater Flow Measurement Systems", Caldon Engineering Report ER-262, available on the Caldon Website.

(13)

~0. ~fZ0:

CALDON LEFM NUCLEAR USER 'S GROUP A GENDA TUESDA Y, FEBR UAR Y 5, 2 002 Tab Time Description Speaker 1

7:45 - 8:00 Registration and Continental Breakfast N/A 2

8:00 - 8:15 Welcome & Mission E. Hauser Caldon, Inc.

3 8:15 - 8:45 Update NRC Approvals E. Hauser Current Climate Caldon, Inc.

Review Times Review Schedules 4

8:45 - 9:45 NRC Guidance on MUR Power Uprate Applications B. Horin Winston & Strawn 5

9:45 - 10:15 Lessons Learned 1:

J. Burford Entergy Thermal Power Optimization Entergy Operations, Inc.

(Appendix K Uprates) 6 10:15-10:30 Morning Break N/A 7

10:30 - 11:00 Lessons Learned 2:

M. Winkelblech Comanche Peak Steam Electric. Station TXU 8

11:00-11:30 Lessons Learned 3:

.B.

Kline Beaver Valley Power Station FENOC 9

11:30 - 12:00 Lessons Learned.4:

T. Yudate History of the LEFM Applications in Japan Hitachi 10 12:00-1:00 Lunch N/A 11 1:00 - 2:30 Data Analysis Profile Changes In Situ and Their H. Estrada Effects on LEFM Systems Caldon, Inc.

12 2:30 - 2:45 Afternoon Break N/A 13 2:45 - 3:15 LEFM, and LEFMI + Product Innovations D. Augenstein Caldon, Inc.

14 3:15 -4:15 Other Applications for LEFM Technology:

E. Hauser RCMS, Blowdown, Steam Flow Caldon, Inc.

15 4:15 - 5:45 Field Trip to Caldon N/A 6:30 Boarding Time Dinner & Entertainment on Gateway Clipper "River Belle" (r.

LEFM User's Group Agenda Count on Caldon Page I

~-&

0;

~

0 wm CALDON LEFM NUCLEAR USER 'S GROUP A GENDA WEDNESDAY, FEBRUARY 6, 2002 Tab Time Description Speaker 16 7:45 - 8:00 Registration and Continental Breakfast N/A Hand Out Questionnaires 17 8:00 - 10:15 LEFM Experience Forum Chordal and External A - Quality Assurance J. Whitehead, Caldon B - Maintenance and Reliability J. Regan, Key Tech.

C - User Group Info Sharing M. Ventura, Caldon D - Data Monitoring Demonstration D. Augenstein, Caldon E - Discussion E. Hauser, Caldon F - Questionnaire E. Hauser, Caldon 18 10:15 - 10.:30 Morning Break....

N/A

"__..Collect Questionnaires.

19 10:30 12:0.0

..LEFM Sbssion A

NRC Commitments'..

J. Regan, Key Tech.

B - LEFM Out of Service:...

J. Regan, Key:Tech.

- Remote Monitbr i-g-.-

S. -Corey, Key Tech.

t The value of Remote Moniitrmg; Case histories. (external f-d-J1orai-20...'12:00- 1:00

-Lunch W.

.. '.--NIA

21.

1:00 -2:30 Brainstorming Session on User GrOufi vities

- 'E. Hauser Caldon, Inc.

22 2:30 - 3:00 Closing Remairks.

E. Hauser

... :Caldon, In*.

23 Appendices NRC Reg6ua.

Issue Summary20032-01 Guidance-NI/A.

on the Contentfof Mkieairement Uncertainty Recapture Power Uprate Applications-......

..24 Appendices.

Paper Capture of Brainstorming Session/Flip Charts To be supplied later LEFM User's Group Agenda Count on Caldon Page 2

=7Za U cz:=Cz 0

CALDON LEFM NUCLEAR USER'S GROUP AGENDA SUNDAY, MAY 4, 2003 - REGISTRATION 5:30 - 7:30 WELCOME RECEPTION - HEAVY HORS D'OEUVRES 7:30 - 9:30 Being held in the Windjammer adjacent to the Atrium MONDAY, MAY 5, 2003 - GENERAL SESSION 8:00 - 5:00 Being held in Thomas Point of the Powerhouse - 1'" Floor Tab Time Description 1

7:30 - 8:00 Continental Breakfast - Windjammer - Users and Spouses 2

8:00 - 8:15 Welcome CNUG 2003 Members 3

8:15 - 9:15 Caldon - Ernie Hauser Weld Integrity 4

9:15-10:15 Caldon-Herb Estrada Update of Velocity Profiles (7) 5 10:15-10.30 Morning Break-in room 6

11:00 - 11:30 Utility Speaker - Waterford - Ray Conigliaro Ultrasonic Flowmeter Caldon LEFM CheckPlus 7

11:30 - 12:00 Utility Speaker - Hitachi-Tadahiro Yudate Recent Status of the LEFMApplications in Japan 8

12:00 -1:00 Lunch-Windjammer 9

1:00 - 1:30 Utility Speaker - Peach Bottom - Jason McDaniel Peach Bottom Atomic Power Station - LEFM Per oe

-aice`M-onitoring 10 1:30 - 3:00 Caldon - Don Augenstein Reliability Update 11 3:00 - 3:15 Afternoon Break-in room 12 3:15 - 5:00 Closing Remarks and Q&A's 6:15 Dinner at Chesapeake Bay Beach Club - Meet in the Lobby at 6:.00 l

Agenda CNUG 2003

Lao' C===E 0

CALDON LEFM NUCLEAR USER 'S GROUP A GENDA TUESDAY, MAY 6,2003 - GENERAL SESSION 8:00 - 5:00 Being held in Thomas Point of the Powerhouse - 1t Floor Tab Time Description 13 7:30 - 8:00 Continental Breakfast - Windjammer - Users and Spouses 14 8:00 - 9:15 Key Technologies - Jenny Regan Measurement Uncertainty & Calorimetric Power Uncertainty 15 9:15 - 10:15 Caldon - Herb Estrada Traceability 16 10:15-10:30 Morning Break - in room 17 10:30 - 12:00 Caldon - Don Augenstein LEFM Check and LEFM Checik~lus Product Development 18 12:00 - 1:00 Lunch -Atrium A 19 1:00 - 2:30 Caldon - Herb Estrada Application Topics 20 2:30 -2:45 Afternoon Break-in room 21 2:45 - 4:00 Caldon - Don Augenstein LEFM Check (Plus) Software Development Rev. K & Beyond 22 4:00 - 5:00 Closing Remarks and Q&A's Free Time to Explore the City 6-WEDNESDAY, MAY 7, 2003 - Continental Breakfast Windjammer 7:30 - 8:00 Sailing or Golfnmg Agenda CNUG 2003

CALDON LEFM NUCLEAR USER -S GRO UP A GENDA C

1 2 C'

c" 0

SUNDAY, MAY 23, 2004 - REGISTRATION 6:30 WELCOME RECEPTION - HEAVY HORS D'OEUVRES 7:00 - 10:00 Being held in the Kennedy Room adjacent to the Lobby MONDAY, MAY 24, 2004 - GENERAL SESSION 8:00 - 5:00 Being held in the Press Room Tab Time Description 1

8:00 - 8:30 Ernie Hauser Welcome CNUG 2004 Members 2

8:30 -10:00 Ernie Hauser Anticipated NRC Action and Response 3

10:00 -10:30 Don Augenstein Update of Reliability Review 4

10:45 - 11:30 Ryan Hannas OE Report - Review since last CNUG Ernie Hauser-*

5 11:30 - 12:00 EneHue 5__ _

11:30_-_12:0 NUPICAudit Review 6

1:00 -2:00 Matt Mlhalcin 6______

New.APUDesign: Short Circuit Detection Herb Estrada 7

2:00 - 3:30 Backup Modes of Operation (with break halfiway) 8 3:30 - 5:00 Open Discussion CNUG 2004 Agenda

CALDON LEFM NUCLEAR USER'S GROUP AGENDA c'L

0.

0

=M TUESDAY, MAY 25, 2004 - GENERAL SESSION 8:00 - 12:00 FIELD TRIP TO ALDEN LABS 12:00 - 4:30 Tab Time Description Tadahiro Yudate - Hitachi 9

8:00 - 8:45 Recent Status of the LEFMApplications (Part 1 - Uprating with External Units in Japan Don Asay - Dominion Millstone Case Study Tom Hokemeyer - CP&L 11 9:45 - 10:15 Brunswick Case Study FW Venturi Investigation Using LEFM2000FE 12 10:15 - 10:45 Herb Estrada Uncertainties In Nozzle-Based Feedwater Flow Measurements Tadahiro Yudate - Hitachi 13 10:45 - 11:15 Recent Status of the LEFM Applications (Part 2 NMIJ Testing Report 11:45 -4:30 Dr. Jim Nystrom -Alden Labs Hydraulic Accuracy Dinner at Boston Billiards and Fenway Park Outing Boston Red Sox vs. Oakland Athletics WEDNESDAY, MAY 26, 2004 Tab Time Description Herb Estrada 14 8:00 - 9:00 Measuring Flow on Advanced Gas-Cooled Reactors, the Pebble Bed Modular Reactor (PBMR).

15 9:00 - 9:30 User's Survey 16 9:30 - 10:30 "Predict the Proffie" Winners Announced 17 i0:30 - 11:30 Open Ideas and Topics for CNUG 2005 12:00 - 1:00 Farewell Luncheon 6

CNUG 2004 Agenda

QQ 0 C=

C

'r 227--

]

-Er::*-Xr CNUG 2005 Final Meeting Agenda MONDAY 5/23/05 Meeting Description Author/Company - Speaker Time Tab Welcome to CNUG 2005 Ernie Hauser 8:00-8:30 1

President's Welcome Cal Hastings 8:30-8:45 2

LEFM 101 Herb Estrada 8:45 - 9:30 3

Group Photo 9:30 - 9:45 Morning Break 9:45 - 10:00 Licensing Process Update -

Measurement Uncertainty Recapture Bill Horin/Winston & Strawn 10:00 - 11:00 4

Power Uprate License Amendments Summary of June 18 NRC Report on LEFMs Ernie Hauser 11:00 - 12:00 5

Lunch 12:00 - 1:00 INPO Perspective on Ultrasonic Flowmeter (UFM) Operations Verification and Peer/Self-Bob GambrilINPO 1:00 - 2:00 6

Assessment LEFM EXTERNAL Users Breakout LEFM CHORDAL Users Breakout Verifying Ernie Verifying Chordal Herb 7E External LEFM Hauser 2:00 - 3:00 LEFM Check and Estrada 2:00 - 3:00 7C Systems CheckPlus Systems Afternoon Break 3:00 - 3:15 Free Discussion External 3:15 - 4:00 Free Discussion Chordal 3:15 - 4:00 Wrap Up - Q/A Session

(

QA CNUG 2005 Final Meeting Agenda TUESDAY 5/24/05 Meeting Description Author/Company -

Time Tab

_________________________Speaker Time___Tab_

Palo Verde - Ultrasonic Feedwater Flowmeters -

Ken Porter/APS 8:00-8:45 8

Return to Service Vandellos Report - AND Manel Cambra/ASCO - Herb Supplemental Root Cause Investigation Source of Estrada 8:45 - 9:30 9

Error Corrective Actions Transducer Update Don Augenstein 9:30 - 10:15 10 Morning Break 10:15 - 10:30 APU Updates Don Augenstein 10:30 - 11:15 11 Ensuring That Ultrasonic Flowmeters Live Up To Their Accuracy Requirements Lunch 12:00 - 1:00 Monitoring LEFMs Herb Estrada 1:00 - 2:00 13 Constructing A Best Estimate Of Feedwater Flow Comparisons of Steam Plant Measurements with.Herb Estrada 2:00 - 2:45 14 Chordal LEFMs (Seabrook & River Bend)

I)I_

I Afternoon Break 2:45 - 3:00 MUR BOP Evaluations Robert Field/Sargent &

3:00 - 3:45 15 Lundy 3:00_-3:45_

1 MUR NSSS Scope Fred Maass/Framatome 3:45 - 4:30 16 Wrap Up - Q/A Session (r

L20

==C C

CNUG 2005 Final Meeting Agenda WEDNESDAY 5/25/05 Meeting Description Author/Company -

Tie Tab Speaker TimeTab Reliability Update Leeanne Jozwiak 8:00-8:45 17 History & Future of Japanese Nuclear Industry Tetsuya Takahara/Marubeni 8:45 - 9:30 18 Innovative Practices for the Installation of Vic Ferraro & Marion 930 - 10:15 19 Caldon Leading Edge Flow Meters Freeland/WSI Morning Break 10:15 - 10:30 New Developments Japanese Architecture Ryan Hannas 10:30 - 11:15 20 User Survey All attendees 11:15 - 11:30 21 Open Forum Free Discussion All attendees 11:30 -12:30 Farewell Luncheon 12:30 - 1:30

CNUG 2004 User Attendance Master Sheet Updated 5/20/2004 Company Site First Name Last Name Phone E-Mail APS CP&L Nuclear Dominion Generation Dominion Generation Ent~ergy Nuclear E-ntergy Nuclear Entergy Operations Entergy Operations.

Exelon Generation FirstEnergy.

FirstEnergy_

Fluids Control Instrument FP&L_

FP&L Framatome Technologies Hitachi, Ltd.

Institute of Nuclear Energy Marubeni MPR Associates NMCCO NMIJ NMIJ Sargent & Lundy._.

Sargent &-Lundy__....

SCE&G SNOC S

TVA TXU Winston & Strawn PaloVerde

-Nine Mile 1 Corporate Millstone 3 Surr.y_ & North Anna Grand Gulf Waterford Waterford Waterford Quad Cities 1 & 2 Davis-Besse Seabrook

• Seabrook A,ll JapanInstallations

.Point Beach 1 & 2 Fluid Flow Division Fluid Flow Division~

Susquehanna1&2 1

Summer Fa rley_,_Hut ch, Vogtle_-

Watts Bar omanche Peak1 & 2 Carl Tom Don Jon Ray_

Tim..

Lonnie Jim Bob Mike Kevin Steve Ian Bret Tadahiro

!Chin-Jang Tetsuyay__

Bob Mel Hiroshi Masaki Bob Robert_

Tunglu Bill_

Ray.-_

Jack Mark Bill Landstrom Stathis Hokemeyer AsaEa.

Thomas B ymes Conigliaro Boehm Weber Foster Wharrv Yeager Liu Hale Watters Boman Yudate Chang Takahara Coward Pedersen 623-393-5129 315-349-4601 919-546-2692 860-444-5303 804-273-2205 601-436-2493 504-739-6229 504-464-3468 309-227-2703 419-321-7564 440-280-7517 886-227093389 603-773-7206 603-773-7534-____

434-835-2677 011-0294-23-5395 886-2-813177717 X6092 011-81-3-3214-9020 703-519-0418 920-755-6566 011-81-29-861-4242 011-81-29-861-4242 011 81 29 861 4228 570-542-3947 312-269-3909 302-622-7270 803-345-4389 205-992-6448 423-365-3076 254-897-6277 RonThomas@dom.com JBYRNES@entergy.qcm rconigl@.entergy.com tboehm@nt~ergy.com lweber~entergy.com_

iames.foster~a)exeloncorp.com rkwharry@firstenergycorp.com Don FAsayadom.com clandstr@apsc.com william.stathis@.constellation.com tom.hokemever(f.tonmail.com

..vinli_ flulds.com.tw steve hale(EUPI.comn ian watters@fpt~corn bre~b9_Man@framnatome anp~com ta~dahiro yudate~pisý.hitch.~p pjchang_@iqe gqv.tvy___

takahara@,fus.9o.'p rcoward@mpr.com Mel. Pedersen~cnmcco.com Sato Terao 4.

Takamoto sato-hirosh~i@ag~itL~ojL terao~rialst~g.jp____

m-takamoto~~tgoj rtmiganotta~pplweb.co robert.m.field(c~saraentlundv.com Magnotta Field Wana

• 1-Bell tungiu.wan.g@

entlundy.com whbelL@_scana.com rlherb@southemco.com Herb Bryant Winkelblech mwinkel TXU-com-.____

whorine-winston.com Horin 202-371-5950 Page 1

AttendanceLUst CNUG 2005.

May 22-25, 2005 Company Site First Name Last Name E-Mail ANAV ASCONandellos Ignasi Balazote ibalazote@anacnv.com APS Palo Verde Dan Fisher dfishe0l@apsc.com APS Palo Verde Ken Porter Ken.Porter@apsc.com Constellation Generation Group Corporate Mike McMahon michael.s.mcmahon@constellation.com Duke Energy Catawba Ralph Neigenfind rgneigen@duke-energy.com Duke Energy Catawba David Wilson dawilson@duke-energy.com Duke Energy Catawba Turner Wood wtwood@duke-energy.com Duke Energy Oconee Lesley Burns Ipbums@duke-energy.com Duke Energy Oconee Steve Hobbs sdhobbs@duke-energy.com Duke Energy Oconee William Rostron wcrostro@duke-energy.com Duke Energy Oconee Bluford Jones gbjones@duke-energy.com Entergy Grand Gulf Jon Byrnes jbymes@entergy.com Exelon Peach Bottom Jim Zardus jim.zardus@exeloncorp.com FirstEnergy Davis-Besse Brian Young bdyoung@firstenergycorp.com Fluids Control Fluids Control Kevin Uu Kevinlu@flulds.com.tw Framatome Frarnatorne Fred Masse Fred.Maass@framatome-anp.com INER INER Yeong-Jen Su yjsu@iner.gov.tw INER INER Chin-Jang Chang cjchang@iner.gov.tw INPO INPO Bob Gambdill gambrillra@inpo.org Marubeni Utility Services Japan Installations Tetsu Takahara takahara@mus.co.jp NMIJ Liquid Flow Standard Norlyuld Furulchl furuichi.nodyuki@aist.go.jp NMIJ Liquid Flow Standard Hiroshi Sato sato-hiroshi@aist.go.jp Petrobras Petrobras Sergio Figuelredo sfigueiredo0l @petrobras.com.br PKMJ Technical Services PKMJ Curt Ciocca cioccac@pkrmj.com PP&L, Inc.

Susquehanna Bob Magnotta rtmagnotta@ppiweb.com Progress Energy H.B. Robinson Rick Harrold dck.harrold@pgnmail.com Progress Energy H.B. Robinson Rich Supler richard.supler@pgnmail.com Sargent & Lundy Sargent & Lundy Tunglu Wang tunglu.wang@sargentlundy.com Sargent & Lundy Sargent & Lundy Robert Field robert.m.field@sargentlundy.com SCE&G VC Summer Bill Bell whbell@scana.com Southern Nuclear Corporate Michael Eidson mgeidson@southemco.com TEPCO Corporate Takashi Sato satoh.takashi@tepco.co.jp TVA Nuclear Watts Bar Jack Bryant jkbryant@tva.gov TXU Comanche Peak Mark Winkelblech mwinkell@TXU.com Welding Services, Inc.

WSI Victor Ferraro vferraro@wsi.aquilex.com Welding Services, Inc.

WSI Marion Freeland mfreeland@wsi.aquilex.com Winston & Strewn W&S Bill Horin whorin@winston.com

I 1~-

first_n.

last-name company email phone Attec.

,Itg Greg Hill AEP, DC Cook gjhill@aep.com 616-697-5134 Yes Kenneth Riches AEP, DC Cook kwriches@aep.com 616-697-5146 Yes Mark Williams AEP, DC Cook mgwilliams@aep.com 616-697-5129 Yes John Simpkins APS, Palo Verde jsimpkin@apsc.com 623-393-5325 Yes Chris Mills CP&L, Brunswick chris.mills@pgnmail.com 910-457*2567 Yes Chuck Baucom CP&L, Robinson chuck.baucom@pgnmail.com 843-857-1253 Yes Tom Hokemeyer CP&LiCorporate tom.hokemeyer@pgnmail.com 919-546-2692 Yes Jim Snelson CP&L/Robinson james.snelson@pgnmail.com 843-857-1129 No Ron Thomas Dominion ronthomas@dom.com 804-273-2205 Yes John Gibson Dominion, Millstone johnj_gibson@dom.com 860-447-1791 Yes Mike Withrow Entergy mwithro@entergy.com 601-437-6247 Yes George Thomas Entergy, Vermont Yankee Transition Teai gthomas@entergy.com 802-451-3072 Yes Jerry Burford Entergy, River Bend fburfor@entergy.com 601-368-5755 Yes Tom Fleischer Entergy, Waterford 3 tfleisc@entergy.com 504-739-6262 No Mike Baker Exelon, Peach Bottom michael.baker@exeloncorp.com 717-456-4094 No Jason McDaniel Exelon, Peach Bottom william.mcdaniel@exeloncorp.com 717-456-4015 Yes Jim Foster Exelon, Quad Cities james.foster@exeloncorp.com 309-227-2000 ext 2703 No Bill Kline FENOC, Beaver Valley klinew@firstenergycorp.com 724-682-5620 Yes Curt Ciocca FENOC, Consultant cioccahouse@worldnet.att.net 724-682-1872 Yes Mark Musulin FENOC, Beaver Valley musulinm@firstenergycorp.com 724-682-5625 No Craig Hengge FirstEnergy, Davis Besse cahengge@firstenergycorp.com 419-321-7898 Yes Mike Yeager FirstEnergy, Perry mJyeager@firstenergycorp.com 440-280-8035 No Mike Rubano FP&L, St. Lucie mikerubano@fpl.com 561-467-7298 Yes LEFM Nuclear User's Group 9:31 AM 1

Ffirsýn, last-name company email phone

Atter, itg Bret Boman Framatome bboman@framatech.com 804-832-2677 No Martin Parece Framatome mparece@framatech.com 804-832-2474 Yes Tadahiro Yudate Hitachi tadahiro yudate@pis.hitachi.co.jp 0294-23-5395 Yes Scott Corey Key Technologies Inc.

scorey@keytechinc.com 410-385-0200 Yes Jenny Regan Key Technologies Inc.

jregan@keytechinc.com 610-274-8258 Yes Tetsuya Takahara Marubeni takahara@mus.co.jp 81-3-3214-9020 Yes Tom McMahon Niagara Mowhawk, Nine Mile mcmahontl @niagaramohawk.com 315-349-4045 No Harv Hanneman NMCCO, Point Beach harv.hanneman@nmcco.com 920-755-7317 Yes Tom Behringer Sargent & Lundy thomas.j.behringer@sargentlundy.co 312-269-7218 Yes m

Frank Calabrese Sargent & L.undy frank.j.calabrese@sargentlundy.com 302-622-7369 Yes Bill Bell SCE&G, VC Summer whbell@scana.com 803-345-4389 Yes Mike Eidson Southern Nuclear Op Co mgeidson@southernco.com 205-992-5978 No Shinichi Kawamura TEPCO kawamura@tepco.com 202-457-0970 Yes Fumihiko Ishibashi Toshiba fumihiko.ishibashi@toshiba.co.jp 650-875-3464 Yes Atsushi Tanaka Toshiba atsushi5.tanaka@toshiba.co.jp 212-596-0614 Yes Stan Nelson TVA Watts Bar Unit 1 sbnelson@tva.gov 423-365-3554 No Jim Swearingen TVA, Sequoyah jdswearingen@tva.gov 423-843-7628 Yes Mark Winkelblech TXU Electric, CP mwinkell@TXU.com 254-897-6277 Yes Bill Horin Winston & Strawn whorin@winston.com 202-371-5950 Yes LEFM Nuclear User's Group 9:31 AM 2

.-ndlng 2003 last-name Ii.-.name Nickname Spouse company email phone Stathis William Bill N/A Constellation Nuclear, 9 Mile william.stathis@nmp.cn.com 315-349-4601 Yes Hokemeyer Thomas Tom Cardo CP&LUCorporate - Progress Energy tom.hokemeyer@pgnmall.com 919-546-2692 Yes Snelson James Jim CP&LlRobinson james.snelson@pgnmail.com 843-857-1129 Yes Thomas Ronald Ron Anne Dominion ron thomas@dom.com 804-273-2205 Yes Waddill John Dominion Yes Zumbo Wendy Wendy Dominion, Millstone WendyEZumbo@dom.com 860-447-1791 Yes Wyspianski Leslaw Les Dominion, Millstone leslawwyspianski@dom.com 860-447-1791 x6800 Yes Thomas Walter Ed Dominion North Anna Power Station EdThomas@dom.com 540-894-2784 Yes Gibson John Jill Dominion, Millstone johnj_gibson@dom.com 860-447-1791 Yes Bymes Jonathon Jon Entergy, Grand Gulf lbymes@entergy~com 601-436-2493 Yes Conigliam Raymond Ray Entergy, Waterford 3 RCONIGL@entergy.com 504.739.6229 Yes Dowhy Thomas Tom Nadine FENOC, Beaver Valley dowhyt@firstenergycorp.com 724-682-7935 Yes Beese Larry Ginger FirstEnergy, Davis Besse 419-321-7543 Yes

_______ ______Torres

________________lwbeese@flrstenergycorp.com 493174 e

Yudate Tadahiro Hitachi tadahiroyudate@pls.hitachi.co.jp 0294-23-5395 Yes Regan Jennifer Jenny Tim Key Technologies Inc.

jregan@keytechinc.com 610-274-8258 Yes Takahara Tetsuya Tetsu Marubeni takahara@mus.co.jp 81-3-3214-9020 Yes Magnotta Robert Bob Lynn PPL Susquehanna LLC rtmagnotta@pplweb.com 570-542-3947 Yes Bieter Walter Walt Sargent & Lundy WALTER.J.BIETER@sargentlundy.com 302-622-7278 Yes Bartoski Thomas Tom Sargent & Lundy Thomas.bartoski@sargentlundy.com 302-622-7275 Yes Eidson Michael Mike Southern Company MGEIDSON@southernco.com 205 992-5978 Yes Horiguchl Masahiro Toshiba International Corp.

masahirol.hodguchi@toshiba.co.jp 212-596-0669 Yes Sakamoto Hiroshi Hiroshi Toshiba International Corp.

himshi6.sakamoto@toshlba.co.jp 212-596-0614 Yes Bryant Jack Tracy TVA, Watts Bar Unit 1 jkbryant2@tva.gov 423-365-3076 Yes Winkelblech Mark TXU Electric, CP mwinkell@TXU.com 254-897-6277 Yes

,Hodn William Bill Winston & Strawn whorin@winston.com 202-371-5950 Yes LEFM Nuclear User's Group 2003 9:38 AM I

October 17, 2003 Michael Baker Program Manager Exelon Nuclear Peach Bottom Atomic Power Station 1848 Lay Road Delta, Pa. 17314-9032 Phone: 717-456-4094

Reference:

Exelon Nuclear P.O. No. 01038929 Caldon, Inc CO-22862

Subject:

Caldon, Inc. Peach Bottom Unit 3 LEFM,/+ System Commissioning Certificate of Compliance Letter Dear Mr. Baker.

Caldon has reviewed the commissioning results, i.e., Fluid Velocity ratios, Sound Velocity ratios, Non Fluid Tau's, Spool Dimensions, Alarm Settings, etc. at the plant operating condition of -98% power and has concluded that the LEFM," + System is operating within its bounding uncertainty of+1- 0.30 % of its rated flow rate. The LEFM/ + System can beused for flow measurement.

Caldon will send a copy of the Field Commissioning Data Package, FCDP-125, by the end of October, 2003.

If you should have any questions and/or comments, please call or email me at emaderae~caldon.net.

Sincerely Yours, Ed Madera Caldon Senior Project Engineer CC: Ernie Hauser, President of Nuclear Division, Caldon, Inc.

Garry Ventura, V.P. of Operations, Caldon, Inc.

Don Augenstein, V.P. of Engineering, Cildon, Inc.